Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2008-04-07

00:00:14Bagderkrz: well, I can tell you that I don't believe in working with a highres version that "compiles" into target versions
00:00:27amiconnNico_P: Happens both with FLAC and MP3 VBR
00:00:37Bagderand I think the people who have done themes for many targets will agree
00:00:52Bagderbut that's all details
00:00:59Nico_Pamiconn: I tested with MP3 VBR... does it also happen to you without shutting down? ie let it stop and resume
00:01:15Nico_PI also start on the last song in the dir for my testing
00:01:17 Quit Jon-Kha (Remote closed the connection)
00:01:29amiconnMaybe it works when playing one album without interruption, but that's not my typical usage case
00:01:52Nico_PI don't think that would change things... are you up-to-date btw?
00:01:56amiconnThe typical resume is not from the start of a song
00:02:00 Join cbr|w [0] (n=cbr@
00:02:44 Quit gregzx ("ChatZilla 0.9.81 [Firefox]")
00:03:48krzBagder: sure it will not completely copy wps, but it will preserve layout as much as possible
00:03:50 Quit desowin ()
00:03:57Nico_Pkrz: I'm also not convinced a unified format for all targets is feasible
00:04:03amiconnD'oh - H180 is still at r1625
00:04:30DerPapsttime to update i'd say :-P
00:04:34Nico_Pamiconn: you need at least r16955 IIRC :)
00:04:38amiconnMeh, current will give me those nasty wrong indices in wps... but I'll try :/
00:04:51kugelr16999 is latest
00:04:57Nico_Pamiconn: oh come on, they only last for a couple seconds ;)
00:04:58kugelI wonder who will go for r17000
00:05:16Nico_Preally? 16999 already? I *haveù to commit!
00:05:27amiconnNico_P: Yes, but for some reason this is most often the time when I check the displayed track info
00:05:46krzNico_P: generally it will preserve layout, this will help to modify wps for new targets
00:05:52kugelNico_P: Gogo!
00:05:54Nico_Pamiconn: I really want to solve it, but I need to find a way that won't break anything else
00:06:20kugelNico_P: Commit a manual change or something but do it !! :P
00:06:22Nico_Pmy plan is to rewrite audio_check_new_track actually
00:07:34 Quit DerDome ("Leaving.")
00:07:51pixelmakrz: I believe in "doing everything explicitely for one screen (resolution, colour depth". I'm pretty sure that even "preserving" a layout would work. Do you know all the different display sizes, colour depth, screen orientation of players Rockbox runs on (not to forget remote screens etc.)?
00:07:57 Quit mcuelenaere ()
00:08:03pixelmaeh, wouldn't
00:09:16Nico_Pkugel: IIRC the discussion about the smooth resize was that it could be made simpler/smaller
00:09:38kugelNico_P: But then having it in the core?
00:09:48Nico_Pbut let's face it, noone has tried to and I don't see it happenning with rotting in the tracker
00:09:52*pixelma counts 14 cabbiev2 versions...
00:09:53amiconnNico_P: I've now updated the H180, and have an album playing. Will report back when it ends
00:09:58 Join gromit`` [0] (
00:09:59Nico_Pamiconn: thanks
00:10:04amiconnThis one is FLAC, btw
00:10:09kugelNico_P: Mainly since the creator retired from contributing
00:10:24Nico_Pkugel: well he merely ported the algorithm
00:10:56kugelNico_P: I just had the idea the remove those cases for scaling in 1 direction, and take only the part where aspect ratio is kept
00:11:04Nico_PI think it can't hurt to have it in the pluginlib, and if/when we decide to move it into the core, we can optimise it then
00:11:53 Join Vampiryan [0] (
00:12:18Vampiryanhi all
00:12:36Nico_PI think I'll leave all special cases in for now
00:12:52Nico_PjhMikeS: got r17000!
00:12:54kugelI wonder a) why simple resize isn't in the core and b) if someone compared the performance of smooth resize against simple resize
00:13:13kugeljhMikeS: Weeehh!
00:13:46moosamiconn had the 16000th IIRC :)
00:13:50*jhMikeS looks forward to 20000
00:14:00bluebrothera general rule is that thing that could be done in advance should be done in advance. Like decoding images to bmp or scaling them.
00:14:13bluebrotheror rendering fonts, ...
00:14:25*amiconn got I08: PrivVio on iriver
00:14:52*amiconn wonders how that's possible as we're running in supervisor mode...
00:14:58Nico_Pjust with audio playack?
00:15:19BigBambiHowever with bmp resize it would a) allow you to use any WPS without needing a PC to resize your entire collection, and b) having multiple versions of every cover
00:15:25amiconnNo, it's something in my remote lcd driver code. I'm reworking it because of the reported failures
00:16:35 Part Spex
00:17:09jhMikeSamiconn: the #0x2700 in the remote drivers should really be #0x2500 since audio DMA isn't normally masked.
00:17:58 Quit bluebrother ("leaving")
00:18:07 Quit gromit` (Read error: 110 (Connection timed out))
00:19:03 Quit CyBergRind|w (Read error: 110 (Connection timed out))
00:19:24peturjhMikeS: any hints what could be causing my crashes in thread.c in the h300 bootloader? If I add a sleep() in iriver_h300.c on line 302 (before usb_init) I get AddrErr and IllInstr
00:20:39 Join tomcat_ha [0] (n=F14_fana@
00:20:40 Part tomcat_ha
00:20:42 Join tomcat_ha [0] (n=F14_fana@
00:21:22jhMikeSis there anything inclined to use something too early? any task switching calls will also enable interrupts.
00:21:32amiconncopy&paste error...
00:21:49 Quit simonrvn ("bbl")
00:22:00 Join TaylorKillian [0] (
00:22:10peturthis is well after interrupts are enabled
00:22:25TaylorKillianhello all
00:22:49TaylorKillianjust wanted to let you guys know that i updated my application (ARM Emulator)
00:22:51jhMikeSpetur: I'll have a closer look.
00:23:04TaylorKilliananyways, let me know if you have any further questions
00:23:20tomcat_haso my itunes cant see my ipod 5.5 gen at all
00:23:33tomcat_haso i cant use the restore option to try to get my ipod working again
00:24:08tomcat_haany other way to restore my 5.5 gen ipod? on windows home media center edition
00:24:11 Join webtaz [0] (n=4fd48d9e@gateway/web/cgi-irc/
00:25:26tomcat_hai also tried dskchk
00:25:32jhMikeSpetur: weird since there is a big long sleep way before it
00:25:39 Quit lee-qid ("aufwiederbyebientotsayonara")
00:25:44 Quit krz (" ?")
00:25:44peturjhMikeS: yes
00:26:09 Quit robin0800 (Read error: 104 (Connection reset by peer))
00:26:26gevaertsTaylorKillian: thanks for letting us know :)
00:26:37peturjhMikeS: it has nothing to do with the sleep() itself, but with timing. If I add printf() in between I can also get the crashes. Or it just plain hangs.
00:26:49 Join andrew_ [0] (
00:27:04 Join Nate101 [0] (n=465c9f89@gateway/web/cgi-irc/
00:27:18Nate101i have a question about videos on my ipod
00:27:28peturI feel there is some bug there that for some reason only happens on my h380 :/
00:27:54 Quit dabujo (Read error: 104 (Connection reset by peer))
00:28:30*jhMikeS wonders about system_init and kernel_init being so late in the game
00:28:30Nate101some movie files i have when i try and sync to my ipod video (30gb) are invalid types but windows media player will play them? what are valid types..
00:28:50BigBambiMPEG 1/2
00:29:07BigBambiat the correct resolution etc
00:29:42Gartralin short: rockbox is picky
00:30:00gevaertsGartral: you're welcome to implement more codecs
00:30:03BigBambiIt only plays the formats it has a decoder for
00:30:20Gartralat a designated resolution
00:30:21jhMikeSpetur: if any of that could touch threading in any way there could be problems
00:30:30BigBambiAnd at the correct resolution, as it is an MP3 player, not a PC
00:30:53Nate101wouldnt it just resize the resolution tho?
00:30:56peturjhMikeS: any of what?
00:31:02BigBambiVirtually all the targets cannot resize a video on the fly, it just isn't powerful enough
00:31:26jhMikeSpetur: any of the code before system_init/kernel_init
00:31:31Nate101oh...i was wondering cuz for some reason it can do it for my movie
00:31:47BigBambiNo, it can't
00:31:57BigBambiRockbox does not resize movies
00:32:07*petur feels unsafe about trying a bootloader with code changes that early. Must ask LinusN :/
00:32:09Nate101no i mean on my regular ipod, im trying to install rockbox now
00:32:24BigBambiDo you have a Video IPod?
00:32:42BigBambiIt has an additional video decoder chip that we don't know how to use
00:32:44Nate101i just got it from my sister, we traded cuz she doesnt like hers and has never used it
00:33:00BigBambiBut in many cases, itunes does stuff on the way over
00:33:01Nate101oh alright, but i can still use rockbox right?
00:33:04amiconnjhMikeS: Late? They happen as the first 2 things in init() (just with the coldstart detection on SH before it)
00:33:13Nate101ok cool cuz i had ipod nano 2nd gen and i couldnt before
00:33:15 Quit webtaz ("CGI:IRC (Ping timeout)")
00:33:31Nate101also before i couldnt find the bootloader, where is it?
00:33:37BigBambiNate101: But with the Ipod video only, you are probably better off watching vids in the Apple firmware, and doing everything else in Rockbox
00:33:49BigBambiNate101: Full instructions are in the manual, please read it
00:34:47Vampiryangood night
00:34:54jhMikeSamiconn: the iriver bootloader does a bunch of stuff before it
00:35:10 Quit waldo ("Konversation terminated!")
00:35:21amiconnAH, bootloader
00:35:49Nate101so i can do everything in rockbox and still watch videos off apple firmware?
00:35:50 Join webtaz [0] (n=4fd48d8f@gateway/web/cgi-irc/
00:36:07jhMikeSIf it's guaranteed to not touch the kernel stuff then it's safe.
00:36:19BigBambiNate101: Yes, the manual tells you how to dual-boot
00:36:31jhMikeSpetur: Is only the H380 doing this?
00:36:31amiconnWell, the ON button detection needs to be done as soon as possible, hence it's above.
00:36:57peturjhMikeS: yes, with my 40GB disk it boots fine
00:37:36peturbut the 80GB disk attached to my h320 also boots fine
00:37:59amiconnIt's done "by hand", i.e. shouldn't influence any kernel stuff. The rtc check is probably necessary too.
00:38:45Nate101do i need to install a theme now or can that be done at a later date
00:39:05BigBambiWhenever you like, but make sure to install the font pack
00:39:28Nate101ok and i can install games later too?
00:39:57jhMikeSpetur: Hard to say at this other than something trashing the thread structures later which produces very quick crashes.
00:39:59 Quit BlakeJohnson86 (Remote closed the connection)
00:40:05BigBambiMost plugins come with - the only ones that need extra installation are doom wads, and various roms for emulators
00:40:08amiconnNico_P: red!
00:40:09kugelNico_P: Well done :)
00:40:27Nate101it says it cant find my ipod
00:40:34Nico_PI'm so stupid
00:41:00markunNico_P: do you want to use it in the jpeg viewer?
00:41:01BigBambiNate101: Have you connected it in disk mode via itunes?
00:41:09Nico_Pmarkun: would be nice
00:41:10peturjhMikeS: one of the crashes was an AddrErr in struct core_entry cores[NUM_CORES]
00:41:26*jhMikeS has image scaling filters which are tiny
00:41:32Nate101ahhh nope
00:42:14Nico_PjhMikeS: I know and hope we can do better, but I've been meaning to commit this for a long time and didn't want to let it rot in the tracker any more
00:43:39 Quit midkay ("Leaving")
00:43:39jhMikeSpetur: I'm fairly certain that problem would have to come later from a trashing of the bss section.
00:44:16pixelmaNico_P: hmm... since when do blank lines above the %wd tag count in the WPS?
00:44:38Nico_Ppixelma: haven't they always counted?
00:44:40kugelNico_P: Why didn't you convert pf also to smooth reisze in that run?
00:44:41 Join crzyboyster [0] (n=6210f00c@gateway/web/cgi-irc/
00:44:50*kugel has 1 patch less in his build, nice
00:44:59Nate101alright i think i connected it in disk mode but honestly i have no lue
00:45:01pixelmaNico_P: I'm not sure but I think not...
00:45:02Nico_Pkugel: because I'm not sure it's fast enough but I'll sure test
00:45:02*jhMikeS looks for the little gaussian and other filter code
00:45:09crzyboysterWhat exactly was r17001 >;revision=17001 ? Is it an album art resize patch?
00:45:16BigBambinot yet
00:45:17tomcat_haNate101: i have a similair problem..... but i already seemingly tried everything and am now looking at getting apple support.
00:45:22 Quit webtaz ("CGI:IRC (Ping timeout)")
00:45:27kugelNico_P: Let me get on my linux box, I'll test
00:45:27Nate101yayy i got it working
00:45:32BigBambitomcat_ha: No, you have a different problem
00:45:39 Quit kugel ("ChatZilla 0.9.81 [Firefox 3.0b5/2008032620]")
00:45:39Nate101ya i have no idea how i did it but now i see rockbox on my ipod
00:45:53crzyboysterSo does anyone know?
00:46:07BigBambicrzyboyster: I already answered you
00:46:28 Part Gartral
00:46:31 Quit crzyboyster (Client Quit)
00:46:55 Quit ompaul (Client Quit)
00:47:14Nate101is it supposed to be like stuck at the do not disconnect screen?
00:47:47amiconnjhMikeS: Urgh, C++ and floats...
00:48:03pixelmaNico_P: probably not important, was just surprised a bit
00:48:06BigBambiNate101: Have you safely removed it?
00:48:15Nate101uhh nope
00:48:33peturhmmm does the h1x0 have a setting to disable the optical link?
00:48:35 Join midkay [0] (n=midkay@rockbox/developer/midkay)
00:48:45BigBambipetur: In rockbox, yes
00:48:58BigBambiIf you mean optical output on/off :)
00:49:02pixelmaNico_P: (because I *believed* it was different, but oh well)
00:49:17peturBigBambi: yes
00:49:26BigBambithen yes :)
00:49:44Nico_Ppixelma: I don't remember any requirement for the %wd tag to be on top on the WPS
00:49:46Nate101ok now im stuck at a screen the says ok to disconnect
00:49:57BigBambiAnd have you disconnected it?
00:50:19BigBambiAnd removed the cable?
00:50:35 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
00:50:45BigBambiMenu + select will reset it
00:51:41 Join TaylorKillia1 [0] (
00:52:01Nate101ok now i get a menu says
00:52:05 Join kugel [0] (n=kugel@unaffiliated/kugel)
00:52:10 Quit ender` (" Did you know that when you smile at someone you exercise nine different muscles and if you beat them to death you exercise")
00:52:17 Quit TaylorKillian (Read error: 104 (Connection reset by peer))
00:52:25 Nick TaylorKillia1 is now known as TaylorKillian (
00:52:37Nate101rockbox boot loder version 2.0....etc....cant load rockbox.ipod file not found hold menu + select to reboot then select+play for disk mode
00:52:39tomcat_hai seeimly lost my recipt and all
00:52:49jhMikeSamiconn: well, it needs converting of course
00:52:58BigBambiNate101: So you haven't extracted a build properly to your iPod
00:53:07tomcat_hai wonder how i can convince i am still covered under warrenty with my online backing thing :X
00:53:08 Quit hannesd (Read error: 113 (No route to host))
00:53:23amiconnjhMikeS: We'd also need an algorithm that can resize on-the-fly, with only little more than 1 line worth of pixels buffered
00:53:24Nate101should i reconnect to the computer?
00:53:33Nico_Pcrap I'm being sloppy
00:53:46amiconnI mean on-the-fly when loading a .bmp
00:53:47kugelNico_P: Is there an easy way to have both pf versions on the target? like renaming the .rock?
00:53:57markunNate101: just unzip the rockbox build directly to you player. It has a .rockbox/ folder with the rockbox.ipod file inside
00:54:03Nico_Pkugel: renaming the rock will work
00:54:04jhMikeSamiconn: then vertical interpolation is severely limited when minimizing
00:54:07 Join simonrvn [0] (i=simon@unaffiliated/simonrvn)
00:54:13BigBambiNate101: Read the installing the build part of the manual again
00:54:18peturBigBambi: care to tell me where the setting for the optical on/off is?
00:54:29BigBambipetur: One mo, I'll fetch the H100
00:54:34saratogaNico_P: why does that bilinear resize code need seperate cases for horizontal, vertical and horizontal + vertical scaling?
00:55:05Nico_Psaratoga: it's optimisation, but as I said in the commit message they can be removed if/when size matters
00:55:31BigBambipetur: Settings -> General Settings −−> Playback −−> Optical Output
00:55:34amiconnjhMikeS: Well, the scaling factor is known in advance, and I do think it should be possible with the help of the destination buffer and the single-line normalized input buffer
00:55:35saratogaah thats good then
00:55:45peturBigBambi: thnx
00:55:48saratogaif we removed those other cases, the scaling code would be quite small
00:56:07Nate101i dont see anything i skipped
00:56:13 Join BlakeJohnson86 [0] (
00:56:21 Quit kugel (Remote closed the connection)
00:56:24BigBambiNate101: So you extracted to the iPod?
00:56:41tomcat_hais the concepttronic grab'n go a worthwhile mp3 player?
00:56:48Nate101if u mean run the installer thing, find the right place, and click complete installation then yes
00:57:27peturBigBambi: So I wasn't dreaming (this setting is missing in the manual)
00:57:30BigBambiNate101: Rerun the installer, go to the second tab and select install build
00:57:42BigBambipetur: Ah, I haven't checked the manual :)
00:58:19 Join kugel [0] (n=kugel@unaffiliated/kugel)
00:58:29pixelmapetur: you'll probably find other settins missing in the manual too (though I'm a bit surprised that such an old thing is missing too) :\
00:58:38*jhMikeS would still prefer to use C++ for things like mpegplayer where abstraction and containment pays off bigtime
00:58:40pixelmasettings too
00:58:55saratogais anyone here familar with variable length coding?
00:58:58Nate101it doesnt say install build it says install bootloader
00:59:08BigBambiAnd what is the second button?
00:59:10saratogai've found yet another bug in that part of the wma decoder, and it'd be helpful to understand how it works
00:59:23saratogaor rather huffman coding i guess
00:59:27Nate101ahahaha thats sneaky
00:59:49BigBambiNate101: Complete install should work, but there was a bug in one of the releases of it
01:00:03jhMikeSisn't huffmann variable-length coding by definition (using prefix codes)? :)
01:00:54saratogayes they're the same thing
01:01:08*Nico_P test builds this time
01:01:13saratogawell more or less, i think its a DFT vs. FFT thing
01:01:40Nate101i still get the ok to disconnect thing
01:01:54 Quit Rob2222 ()
01:02:14saratogabasically i'm wondering what changing the depth of the binary tree actually does
01:02:34 Quit kugel (Client Quit)
01:02:40jhMikeSlimits the maximum code length
01:03:31jhMikeSor changes it since the maximum code length should be equal to the maximum tree depth
01:03:47 Quit Soap ("Leaving")
01:04:11jhMikeSdepth, path length from root to leaf or whatever :p
01:04:22Nate101ya i downloaded and installed and everything but i keep getting the same errors
01:04:23tomcat_haSo i heared that there are more mp3 players then ipod classic that have 160 gb of memory
01:04:31tomcat_haany idea where i can find them?
01:04:45BigBambitomcat_ha: Google? This channel is for Rockbox only
01:04:45 Join Llorean [0] (
01:05:16saratogajhMikeS: oddly enough, the ffmpeg recommends changing that to conserve memory at the cost of CPU time
01:05:22saratogawhich confuses me
01:05:45saratogai would expect that the depth would be determined by the file being decoded
01:05:53amiconnNico_P: Resume after playlist end is still possible (only once though - next try will splash "Nothing to resume")
01:06:05amiconnI got the last 3 seconds of the last song
01:07:31*amiconn tries something
01:07:59jhMikeSsaratoga: so somehow they allow an increase in path length to conserve memory which I guess makes it approach O(n) instead of O(log(n) or something. I'm just guessing since I'm not actually looking at the code itself.
01:08:27Nate101can anybody help me
01:08:43BigBambiOnly if you ask a question
01:08:50 Join soap [50] (n=soap@rockbox/staff/soap)
01:09:03Nate101i still can't get that screen off, and ive tried installing both ways like you said
01:09:19BigBambiWhich screen off?
01:09:40Nate101the ok to disconnect screen
01:09:41Nate101even though it is disconnected
01:09:54BigBambiSo reset it again like last time with menu + select
01:10:14 Join kugel [0] (i=kugel@unaffiliated/kugel)
01:10:16Nate101i did
01:10:32jhMikeSIt's an array bintree?
01:10:36Nate101i get that error message say ing it cant load rockbox.ipod
01:10:47BigBambiSo it still isn't there properly
01:10:57BigBambiCan you try to extract it manually?
01:11:09Nate101hmm i can try
01:11:28BigBambiConnect to the computer, download the correct build from, then extract it directly to your ipod
01:12:04kugelNico_P: I can't say that it's slower on my e200, but I also can't say that it looks better
01:12:27Nate101how do i extract it directly to my ipod
01:12:33 Join Rob2222 [0] (
01:12:41BigBambiNate101: Do you know how to extract a zip?
01:12:58BigBambiThen do that to your ipod
01:13:11BigBambiIf it is H: then unzip to H:\
01:13:30*jhMikeS must point out that pegbox fails to build on gigabeat S
01:14:05amiconnPegbox has weird controls on many targets
01:14:11***Saving seen data "./dancer.seen"
01:14:58 Quit BigBambi (Remote closed the connection)
01:15:45 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
01:16:32jhMikeSno keymap is defined but it doesn't say "Unsupported Keymap" but instead complains about each button.
01:17:08 Quit moos ("Rockbox rules the DAP world")
01:17:55*jhMikeS should just toss a keymap into it
01:18:29kugelNico_P: What gets acutally resized in pf? The cached bmp only, or the zoom too?
01:19:05kugelresized with simple_resize I mean
01:19:35BigBambipixelma: Thank you for the cabbie v2 changes on grayscale btw. I can actually read text over the logo on the background now :)
01:20:04Nico_Pkugel: IIRC it's the bitmap after it's loaded
01:20:11kugelAhh Nico_P, I see, pictureflow uses another resize for the zoom
01:20:17pixelmaBigBambi: thanks - and me too :)
01:20:21kugelThat's why I can't spot a difference
01:20:21Nico_PI think it's written to a cache file after being resized
01:20:49Nico_Pthe improvement will be visible if you use 100% zoom
01:21:04kugelShould it be possible to use simple/smooth for the zoom as well?
01:21:11kugelNico_P: Hardly, with 100x100 album arts ;)
01:21:31Nico_Pno, the zoom is comptletely different code
01:22:15kugelI see, but it could be changed, can't it?
01:22:31 Quit BHSPitLappy (Remote closed the connection)
01:23:26kugelHmm, make a test 150x150 album art, let's see
01:23:50amiconnNico_P: Quickest way to reproduce the resume bug: Start playing the second last track of a directory. Stop somewhere within that track. Resume, and let it play up to the end of the dir.
01:23:54Nico_Pkugel: it won't be easy to change
01:24:19Nico_Pamiconn: that's not exactly what I call quick :/
01:24:29amiconnIt's important to start on the second last (or earlier) track, and to stop somewhere within the track. Last track won't cause it, as well as starting at a track boundary
01:24:40Nico_Pah, "second last". I read "second"
01:25:17Nico_PI'll try
01:25:22amiconnIt *may* be important to use something vbr - not sure. Most of my music is vbr
01:25:32saratogaMaurus Cuelenaere
01:25:36saratogaMaurus Cuelenaere
01:25:41DBUGEnqueued KICK saratoga
01:25:41saratogaMaurus Cuelenaere
01:25:48amiconnsaratoga: ?
01:25:49Nico_Psaratoga: yes?
01:25:54saratogabroekn paste
01:26:00saratogain the webclient
01:26:15saratogai was trying to paste his name into the client and it pasted it into the chat instead
01:26:22kugelNico_P: I guess if only the cached bmp's are resized, there won't be any performance impact, except maybe a slightly higher cache build time
01:26:33saratogaregardless, has there been any thought to giving that guy whos name i cannot copy and paste svn access
01:26:49saratogahis CVM patch is already > 200kb
01:27:08saratogamaybe parts of it should start going into the svn before they become unmanageable
01:27:59saratogaplus i think parts of it would be useful for the M:Robe port
01:28:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
01:28:21*Nico_P is in favor
01:29:15pixelmasaratoga: which M:Robe port? ;)
01:29:55Nate101is there anyway i can manually install the bootloader?
01:30:09BigBambiyes, the manual tells you how
01:30:11Lloreansaratoga: Aren't there some things he really needs to clean up in his patch (last time i heard, at least) or has that been addressed?
01:30:45BigBambiNate101: But if it is complaining that it can't find rockbox.ipod then the bootloader is there
01:30:53Nate101no its not anymore
01:31:02Nate101im trying to reinstall the bootloader
01:31:07kugelNico_P: My 150x150 AA doesn't get resized in both cases..
01:31:09Nate101cuz i accidentally clicked uninstall bootloader
01:31:12saratogaLlorean: i think they've mostly been addressed
01:31:16Nate101but when i do it says can't open ipod
01:31:26saratogaand anyway, if hes just adding stuff to his part of the target tree ...
01:31:47Lloreansaratoga: It still needs to meet the coding guidelines...
01:32:00*Llorean doesn't know what was wrong with it before.
01:32:11LloreanI just remember that there was some sort of list of things that he needed to fix about it.
01:33:14 Part DavidSG
01:33:24*Nico_P sees some tabs and some very long lines
01:34:15Nico_Pbut I still thinks he would be better off in SVN
01:34:52Nico_Pamiconn: I got the 3 sec resume
01:35:00 Quit mrfree (Read error: 110 (Connection timed out))
01:35:11LloreanNico_P: Couldn't we just say "we'd like you to clean up your patch so that we can give you access and have you commit it?"
01:35:43*petur gives up on the bootloader for now and goes to sleep
01:35:46 Quit petur ("Zzzzz")
01:35:47Nico_PLlorean: yes. we should also mention that it would be better if he could break it down in logical pieces if possible
01:36:21 Quit Nate101 ("CGI:IRC (Ping timeout)")
01:39:37 Quit ali_as_ ("zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZz")
01:39:48MassaNico_P, kugel, amiconn, linuxstb, disorganizer: I just added the viewport bgcolor patch to the flyspray tracker as FS #8856 :)
01:39:57soap is funny, but shouldn't we delete it?
01:40:12soapI'm not going to stick my neck out to do it, though. ;)
01:40:46MassaComments are (of course) welcome!
01:41:19*pixelma doesn't want the wma debug output in the sim console to not overwrite the complete backlog :\
01:41:57*gevaerts tries to parse pixelma's sentence and wonders if it means what she wants it to mean :)
01:42:11Lloreansoap: I really don't want to, but I suppose it should be.
01:42:18pixelmagevaerts: noticed
01:42:30*gevaerts was just too late
01:42:35 Quit advcomp2019 (Read error: 104 (Connection reset by peer))
01:42:41DerPapstDEBUGF them :-P
01:42:48 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
01:42:54soapthat is exactly how I feel, Llorean.
01:42:58*Massa just wants to go to bed and sleep (local time is 01:40 and I have to stand up in a few hours ;)
01:43:06Lloreansoap: But I did it anyway.
01:43:09pixelmaother prople can read the forums at this moment? more :\
01:43:33DerPapstGood night all :-)
01:43:40saratogaLlorean: no has commented on his patches in FS #8686 so far
01:44:03saratogaand he did break up a few of the pieces a while back but no one commented on them
01:44:19kugelMassa: Nice
01:44:31saratogamaybe someone familar with adding a port could post a couple lines in there explaining what we expect?
01:44:33kugelMassa: Same here, should go to bet too :(
01:44:34 Quit DerPapst ("So Long And Thanks For All The Fish!")
01:44:35LloreanMassa: I don't understand why you're commenting on overlapping viewports if it's not allowed?
01:44:45saratogai'm hesitant to do it myself
01:44:52pixelmakugel: I bet...
01:45:02Massakugel: so now you can have a look at ;)
01:45:12kugelI allready looked :)
01:45:34Lloreansaratoga: Well one expectation is that it meets the coding guidelines, since we don't want to commit "new" (non-ported) code that doesn't meet them. I don't know what else. If I recall, there was code that wasn't organized properly (edits to files when they should be new files) but I don't know about that.
01:45:37MassaLlorean: well, it was a discussion earlier here - and I thought I add it as note. Shall I remove that note?
01:45:59LloreanMassa: Well I'm more curious, you describe what it does when viewports overlap. Does it have code for this case, or is this just a side effect, or what/
01:46:00kugelNot sure why you put the *parse_viewport_opt_param in an extra function though
01:46:03saratogaLlorean: do you remember who had looked at it before?
01:46:07kugelIs it meant to get reused?
01:46:23Lloreansaratoga: No clue. Sorry
01:47:06MassaLlorean: actually it's just the same behaviour as is atm (the backdrop colour will always be used) - there's no overlap handling...
01:47:55Massakugel: that's just preparation for later usage in parse_list (when it's able to handle optional parameters) ;)
01:47:56Nico_Psaratoga: Bagder committed part of it
01:48:14saratogaNico_P: whats the rev number?
01:48:27 Quit spiorf (Remote closed the connection)
01:48:39kugelMassa: IMO the || case should be left out
01:48:49saratogaah r16347
01:48:54*jhMikeS wonders why lists with a hidden seletion are so misbehaved at scrolling
01:49:00kugelbut it will probably be needed when more optional parameters come
01:49:24Massakugel: and to make it more readable, btw. shouldn't I used "inline" for that method?
01:49:29saratogaall before FS #8686 was open though
01:50:25Massakugel: why shall || be left out? (to be honest I thought about adding that also to the bg_pattern parameter...)
01:50:33 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
01:50:36 Quit gevaerts ("GSoC mentors : please look at this task list...")
01:51:06kugelAs I said, it will probably be needed if more optional parameters come ;)
01:51:58Massanethertheless - I *now* leave and go to bed - only 4,5 hours left to sleep! Bye everybody!
01:52:30 Quit Massa ("ChatZilla 0.9.81 [Firefox]")
01:52:50 Join BlakeJohnson86 [0] (
01:52:51kugelMassa: Don't you think it was better to make the function return a bool and edit the char over the pointer?
01:53:32kugelI'm gonna follow him I think :P
01:53:38kugelgood night
01:53:50 Quit kugel ("ChatZilla 0.9.81 [Firefox 3.0b5/2008032620]")
02:03:23Nico_Pamiconn: for some reason, playlist_end isn't true when audio_stop_playback() is entered
02:06:54soapok, since-4weeks,html only shows through 16999, and the front page only shows back to 17002, so is one of those missing two the commit of bmp resize?
02:07:40soapnevermind - I can look at the SVN mailinglist. :)
02:08:13saratogaits r17001
02:08:14pixelmait's the red commit in the build table ;)
02:08:48soapstupid question follows: Does this work in WPS, or (it appears) only in pluggins?
02:09:06saratogaits in the plugin library so i assume plugin only
02:09:31saratogathough eventually it'll probably go core i guess
02:09:43Nico_Psoap: as saratoga said: no change in the WPS
02:09:47saratogabilinear interpolation is extremely simple, so it shouldn't use much binsize at all
02:10:19pixelmaNico_P: will grescale eventually get it too?
02:10:29Nico_Pmy thinking for committing this was that it would probably generate interest into getting a better version into the core
02:10:48Nico_Ppixelma: hopefully but I don't see myself implementing it for now
02:13:20pixelmagreyscale too
02:13:55*pixelma should give up typing completely for now and get some sleep too, night
02:15:09 Part pixelma
02:22:14Nico_Pamiconn: I traced the
02:22:55Nico_Poops. here's the end: ...line responsible for this back to r10039
02:32:14Nico_PI need to investigate the fix more
02:32:17Nico_Pbut now, bed time
02:34:04 Quit Zarggg ()
02:35:57 Quit Nico_P (Remote closed the connection)
02:38:53 Join Zarggg [0] (
02:39:25saratogaanyone with a c200 feel like posting the results of a battery bench?
02:39:36saratogai'd like to calibrate the c200's battery one of these days
02:40:57 Quit amiconn (Nick collision from services.)
02:41:04 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
03:14:15***Saving seen data "./dancer.seen"
03:16:59 Nick fxb is now known as fxb__ (
03:16:59 Quit Xerion (Read error: 104 (Connection reset by peer))
03:17:28 Join Xerion [0] (
03:41:53 Join hooncakes [0] (
03:46:58 Quit Thundercloud_ (Remote closed the connection)
03:54:19 Quit hooncakes ()
03:55:28 Quit dandin1 ()
03:59:09 Join cool_walking_ [0] (
04:08:55 Join Shaid [0] (
04:09:08 Join Bagder_ [0] (
04:23:05 Join BlackChaos [0] (
04:23:05 Quit BlackChaos (Client Quit)
04:27:01 Quit Bagder (Read error: 110 (Connection timed out))
04:47:21 Join sarixe [0] (
04:48:52 Quit Shaid (Read error: 110 (Connection timed out))
04:52:49 Quit saratoga ("CGI:IRC (EOF)")
04:58:06 Quit perrikwp (" ajax IRC Client")
04:59:28 Quit jhMikeS (Nick collision from services.)
04:59:34 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
05:04:42 Join CyBergRind|w [0] (n=cbr@
05:06:26 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
05:08:12 Quit Xerion (
05:08:12 Quit BlakeJohnson86 (
05:08:12 Quit cbr|w (
05:08:12 Quit ch4os_ (
05:08:12 Quit joshin (
05:08:12 Quit HEx (
05:08:12 Quit nicktastic (
05:08:12 Quit gibbon_ (
05:08:12 Quit leox (
05:08:12 Quit BrianHV (
05:08:12 Quit lostlogic (
05:08:12 Quit sbhsu (
05:08:12 Quit pabs (
05:08:12 Quit daven (
05:08:37NJoingibbon_ [0] (n=joel@
05:08:37NJoinnicktastic [0] (n=nick@unaffiliated/nicktastic)
05:08:39NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
05:08:58NJoinjoshin [0] (n=joshin@VDSL-130-13-240-229.PHNX.QWEST.NET)
05:09:01NJoinsbhsu [0] (
05:09:04NJoinXerion [0] (
05:11:37NJoinpabs [0] (
05:11:47NJoinBlakeJohnson86 [0] (
05:11:47NJoinHEx [0] (
05:11:47NJoinleox [0] (
05:11:47NJoinBrianHV [0] (
05:11:47NJoindaven [0] (
05:11:50 Join ch4os_ [0] (
05:11:50 Quit BrianHV (Connection reset by peer)
05:11:50 Join BrianHV [0] (n=bhv1@
05:11:51***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
05:14:02 Quit Zarggg ()
05:14:18***Saving seen data "./dancer.seen"
05:14:19 Join Zarggg [0] (
05:17:04 Join DiceR [0] (n=cbceb6be@gateway/web/cgi-irc/
05:18:09DiceRAnybody here using 'black glass' theme with a gigabeat?
05:20:51DiceRTheme seems broken after upgrading to a new build :(
05:22:47 Join ol_schoola [0] (
05:24:52 Quit m0f0x ()
05:26:36 Join codin [0] (
05:34:34 Join DavidSG [0] (
05:34:49 Join wpyh [0] (n=3b42f515@gateway/web/cgi-irc/
05:35:37 Quit Horscht ("")
05:36:10 Part codin ("Leaving")
05:39:35wpyhhi guys
05:39:50wpyhthis is the first time I'm here
05:40:11wpyhI tried to connect to the irc servers but I can't seem to (pidgin says "connecting..."
05:40:25wpyhmaybe because I'm in China :p
05:40:57wpyhso I'm using cgi right now
05:44:55 Quit DiceR ("CGI:IRC")
05:45:00advcomp2019wpyh, you have a question about rockbox
05:52:14wpyhwell, nothing important :p
05:53:12wpyhjust like to hear about the meizu m6 port
05:55:31wpyhand by the way
05:55:32 Quit nplus (Read error: 104 (Connection reset by peer))
05:59:43wpyhi've got a rockchip-based player
05:59:49wpyhhas anyone done a port on such chips?
06:00:00wpyhI mean, I don't see any on the ports list
06:00:09wpyhbug maybe someone has experience
06:03:33DavidSGThey probably will tell you if you want a new port "start it your self and maybe someone will join in", in fact i do believe people ask frequently about new ports. If only they had some location to store the answers to such questions.
06:06:04wpyhI would like to start myself, but... I simply don't have the skills
06:06:17wpyhI would be happy to collaborate though
06:06:24wpyhthe player is OPPO D29
06:09:25DavidSGi just googled it i cannot seem to find any english page about it.
06:10:15 Join Shaid [0] (
06:19:49advcomp2019wpyh, you can look at the "new port" in the forums you can find out
06:21:14 Quit sarixe ("(EE) Failed to load "quit" module")
06:23:56wpyhDavidSG: most likely it's only sold in China
06:25:14wpyhadvcomp2019: There's a topic there but it's old and not updated anymore. Maybe I'll update that topic :)
06:27:24 Quit Zom (Remote closed the connection)
06:27:35 Join Zom [0] (
06:29:55 Join RoC_MasterMind [0] (
06:30:29RoC_MasterMindAny major problems with rockbox on ipod old 3rd gen?
06:31:10 Part DavidSG
06:35:04 Join DavidSG [0] (
06:37:09 Join thegeek_ [0] (
06:41:45 Quit RoC_MasterMind ("Leaving")
06:42:36 Join countrymonkey [0] (n=4b05639a@gateway/web/cgi-irc/
06:45:23countrymonkeyfs#s8712, 8720, 8722, 8788, 8761, and 8834 are in committable states. Could they be applied to svn?
06:45:51 Join webguest35 [0] (n=60e016c4@gateway/web/cgi-irc/
06:46:14webguest35Hey Can someone help me out with my rockboxed ipod video ?
06:47:00countrymonkeywhat's wrong? I am a former rockbox ipod user.
06:47:56webguest35im trying to get it to play Video's
06:48:36countrymonkeywhat format?
06:48:38webguest35and for somereason i dont find it anywhere
06:49:40countrymonkeybleepity bleep. Rockbox cannot play mp4 vids AFAIK. It has mpegplayer but that only plays mpegs video streams 1 and 2.
06:50:54 Join JdGordon|uni [0] (i=82c20d6a@rockbox/developer/JdGordon)
06:51:28webguest35ok pretty much dont understand what you mean by Streams 1 and 2 and AFAIK
06:51:53 Quit Zarggg ()
06:53:03ShaidSimply put: the mpegplayer in rockbox is not capable of playing MP4 videos
06:53:08Shaidto do that, use the original firmware.
06:53:23JdGordon|uniarg.. WTF? playback has gone fubar!
06:53:41JdGordon|uni3 tracks in a row hacve jumped to the next track after about 40sec or so!
06:54:18JdGordon|unioh joy.. its now directory skipped!
06:54:49 Quit thegeek (Read error: 110 (Connection timed out))
06:55:03 Quit XavierGr (Nick collision from services.)
06:55:16 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
06:57:32countrymonkeyplease explain. You sound so out-of-whack.
07:01:41cool_walking_webguest35: read the PluginMpegplayer wiki page.
07:05:44countrymonkeyWhy are there 120 people but half of them I see without fail everytime I come here?
07:06:29countrymonkeyAnd I don
07:06:36countrymonkeyand I don't mean devs
07:09:03cool_walking_Cause I work with computers, so I'm always on one..
07:09:22countrymonkeyI mean on this chanel.
07:09:41cool_walking_'Cause I'm interested in Rockbox
07:10:46countrymonkeyso am I, but I'm only on this thing about 1 hour a week.
07:13:06cool_walking_It doesn't cost anything to just leave the windows here and check it every now and then...
07:14:06countrymonkeyfor some reason, when I do that, the messages stack up way too quickly.
07:14:22***Saving seen data "./dancer.seen"
07:20:08 Quit csc` (Read error: 110 (Connection timed out))
07:23:35 Quit JdGordon|uni (" ajax IRC Client")
07:23:41 Quit webguest35 ("CGI:IRC (EOF)")
07:25:40 Quit countrymonkey ("CGI:IRC (EOF)")
07:30:04 Join countrymonkey [0] (n=4b05639a@gateway/web/cgi-irc/
07:32:45countrymonkeyGosh, I know this only has a vague connection to rockbox, but does anyone know how to get cassettes into an rb playable format? I don't like my cassette player much anymore; it is getting old.
07:33:35 Part toffe82
07:47:54 Quit XavierGr ()
07:48:53 Join DerDome [0] (
07:55:34 Quit countrymonkey ("CGI:IRC (Ping timeout)")
07:57:27 Join Chronon [0] (
07:58:09 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:07:07 Join Rob2223 [0] (
08:08:53 Join Jon-Kha [0] (
08:10:16 Join argumentD [0] (
08:10:25 Quit DerDome (Remote closed the connection)
08:10:38 Join GodEater [0] (i=c2cbc962@rockbox/staff/GodEater)
08:10:58 Join DerDome [0] (
08:14:12 Join Bagderr [0] (n=daniel@gateway/web/cgi-irc/
08:19:54 Quit DerDome ("Leaving.")
08:23:57JdGordonhow do i get stereo 16bit raw PCM from a mp3?
08:25:30 Quit Rob2222 (Read error: 110 (Connection timed out))
08:27:52Bagderrat midnight UTC today the gsoc application period ends
08:28:33Bagderrwe're at 18 applications right now
08:28:52Bagderr(and 5 additional ones marked "Ineligible")
08:37:07DavidSGlast day, eh?
08:40:45 Join davina [0] (
08:41:03Bagderrthe submission rate has practically stopped so I don't expect any more applications now
08:42:30 Part LinusN
08:43:26 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:44:04 Part LinusN
08:44:35 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:47:06GodEaterI'm still not sold on the android proposal. Symbian would be a much better choice in my opinion.
08:47:34GodEaterI'm really not seeing the point of porting Rockbox as an App to a platform which doesn't really exist yet.
08:47:34Bagderryes, but I think he picked Android first and then Rockbox...
08:47:40GodEaterI'm sure he did
08:48:24GodEaterI couldn't really get a feel from the public comments on it how well other people have received the idea though.
08:48:39GodEaterdo you think it's worth doing on Android?
08:48:58Bagderrif it would use a working approach, I would
08:49:38BagderrI mean, not hackish in a sense that could actually run on a typical android phone
08:49:40GodEateris there a timeline for how soon Android devices will become available ?
08:50:18Bagderr"the first handsets built on the platform become available in the second half of 2008"
08:50:27Bagderrtheir site says
08:51:20GodEaterand no doubt those first devices will suck horribly
08:51:47Bagderrand won't be available to mortals but only from weird sources and looking "first gen" all over
08:51:55GodEatervery OpenMoko
08:52:10Bagderrwhich btw I was told very bad things about :-/
08:52:12GodEaterdid I imagine you saying you'd seen one of those in the "flesh" ?
08:52:20Bagderryes I held one
08:52:28Bagderrdead though so I couldn't try the sw
08:52:41GodEaterwas the unit as bad as it looks in pictures ? :)
08:52:56BagderrI talked to a guy who has a dev openmoko and who's kept along with the project
08:53:10GodEaterwhat bad things did he have to say about it ?
08:53:10Bagderrhe basically said it sucks
08:53:23GodEaterspecifics ?
08:53:27Bagderrunstable unstable unstable
08:53:30 Quit argumentD (Read error: 113 (No route to host))
08:54:05Bagderrso even though they've been out for quite some time, you still can't make a solid phone call with it
08:54:28GodEaterthat's such a shame - I was really hoping it would turn out to be decent
08:54:34Bagderrbut about the unit, it feels pretty good although too large
08:55:03amiconnSo a device that's unstable when used for its main purpose.... reminds me of swcodec rockbox a bit :/
08:55:04 Join Neovangl1st [0] (n=Neovangl@
08:55:27Bagderralthough I use swcodec fine almost every day
08:55:43*GodEater does too
08:55:53GodEaterI wouldn't call swcodec "unstable unstable unstable"
08:55:57GodEaterI'd call it quirky :)
08:56:01BagderrI'd agree
08:56:12Bagderrand the quirks hardly ever seem to hit me!
08:56:30GodEater"responds badly to freak outbursts of solar wind"
08:56:47 Quit Neovanglist (Read error: 104 (Connection reset by peer))
08:58:08 Join ender` [0] (
08:58:45Bagderropenmoko is in gsoc2008 but Android isn't
08:59:13amiconnThat's why I said 'a bit'. Some of the quirks hit me more or less regularly, like the wrong playlist index during track change, or the resume after playlist end
08:59:15GodEaternot entirely surprising
08:59:29Bagderr"Optimize Openmoko code on running on 48MB ram devices." ...
08:59:39amiconn...and I know that there are others, not hitting me usually because I don't use the affected features
09:02:01Bagderrindeed, but that makes it far from unusable
09:04:04LinusNit's still quite lousy that the playback is that unstable after all this time
09:04:37Bagderryeah, but that just reflects the nature of this project
09:04:58Bagderrwe're a lot more about the fun of adding new stuff/ports than refining the existing functionality to perfection
09:05:48markunwpyh: which rockchip CPU is inside your player?
09:06:07markunthe older ones are (I thik) mips based, the newer ones ARM.
09:06:09Bagderrhaha, and exactly then my sansa crashed...
09:06:30markunwpyh: and welcome to #rockbox :)
09:09:36 Quit TaylorKillian (Read error: 110 (Connection timed out))
09:11:07amiconnmarkun: Afaik the older rockchips are a Z80 on speed with a dsp tacked on
09:12:09markunah, Z80, I didn't recognize the instruction set
09:12:39Bagderr8bit wonders
09:12:48DavidSGWhat is the location of the WPS parser code?
09:13:00BagderrI guess they may have gotten 16bit extended over time?
09:14:19 Join nikosapi_ [0] (
09:14:25***Saving seen data "./dancer.seen"
09:18:31*DavidSG found what he was looking for
09:18:41 Join petur [50] (n=petur@rockbox/developer/petur)
09:19:27 Quit simonrvn (Read error: 113 (No route to host))
09:19:33 Join simonrvn_ [0] (i=simon@unaffiliated/simonrvn)
09:19:40 Nick simonrvn_ is now known as simonrvn (i=simon@unaffiliated/simonrvn)
09:21:13 Quit nikosapi (Read error: 110 (Connection timed out))
09:32:54 Join TaylorKillian [0] (
09:33:54*LinusN spots a gsoc candidate
09:36:05markunLinusN: which one?
09:36:19LinusNmr Killian
09:36:23markunah :)
09:36:42markunTaylorKillian: which proposal is yours?
09:37:16markunLinusN: which proposal was it? :)
09:37:29LinusNSkyeye emulator
09:38:28TaylorKillianmarkun: I proposed an ARM Emulator, also the itunes database for rockbox
09:38:30LinusNor it could be the anti-stun gun plugin, not sure
09:38:59TaylorKilliandoubting the latter will get accepted based on mentor feedback though
09:39:28 Quit linuxstb (Read error: 104 (Connection reset by peer))
09:39:37TaylorKillianhows it going LinusN
09:41:26 Join J [0] (
09:41:35TaylorKillianheading off to bed, gnight
09:45:30 Join phinze [0] (
09:48:04 Quit Ravendug ()
09:48:32 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
09:51:10wpyhAddicted: hi markun
09:51:16 Quit Shaid (Read error: 110 (Connection timed out))
09:51:20wpyhit's RK2602A
09:52:01wpyhwhy is there an "Addicted:" prefix there?
09:52:45linuxstbDid you press TAB?
09:57:24wpyhoh yeah
09:57:30wpyhactually, I pressed Alt-Tab
09:57:44wpyhbecause I'm using Firefox and I'm switching tabs
09:57:51wpyhAny way to turn this feature off?
09:58:18markunwpyh: don't know, but it's usually a useful feature
09:58:26markundepends on your irc client
09:58:49wpyhI'm using the cgi irc client, inside Firefox
09:59:15wpyhit's because I can't connect using pidgin (maybe because of the Great Firewall here)
09:59:33 Join mrfree [0] (
09:59:48markunamiconn: Are the RK26xx Z80 based? Did you find some info somewhere?
10:00:31markunwpyh: maybe this one is nicer:
10:00:49peturwpyh: try connecting to freenode on port 8000 ;)
10:00:50 Join mib_ldf3iipt [0] (i=8259a0a6@gateway/web/ajax/
10:01:17 Quit mib_ldf3iipt (Client Quit)
10:03:22wpyhmarkun & petur: ok, let's try those
10:04:25 Quit wpyh ("CGI:IRC 0.5.9 (2006/06/06)")
10:05:46 Join wpyh [0] (
10:05:55wpyhhi, I'm back
10:06:04wpyhI'm using port 8000 to connect
10:12:36GodEateranyone else here that bought one of the those cheap c240s recently that's started using it with a microsdhc card ?
10:12:39markunwpyh: I think I'll have time this week to work on the port, and the first test program I want to make should work on all models
10:12:55markunyou have an M3 and M6SP, right?
10:17:54 Join Neovanglist [0] (n=Neovangl@
10:18:18 Quit Neovangl1st (Read error: 104 (Connection reset by peer))
10:19:46 Join gevaerts [0] (
10:20:14markungevaerts: did you by any chance already receive your player?
10:20:19 Join kugel [0] (i=kugel@unaffiliated/kugel)
10:20:30*gevaerts did indeed get a package in the mail today
10:20:47markunopened it up yet? ;)
10:20:56gevaertsI haven't done much with it yet, just booted once
10:21:18wpyhhi, sorry if I'm a bit late at replying
10:21:29markunno problem
10:21:36wpyhmarkun: yes, I have an M3 and M6SP
10:22:02markunwpyh: gevaerts now has an M6SL :)
10:22:26wpyhI'm looking gevaerts: wow
10:22:38wpyhI'm looking for an M6SL here :p
10:22:40wpyhI mean secondhand
10:22:45wpyhsince the new one is too expensive
10:25:19 Join webguest21 [0] (n=d9e1c6cf@gateway/web/cgi-irc/
10:26:05webguest21hi, H140, can I flash to RAM only or ROM only ?, or do I have to flash to both
10:27:41markunwebguest21: I don't really understand the question. What exactly are you trying to do?
10:29:42webguest21I have flashed my player, and alls well, what I dont understand is, Do I have to flash to RAM and ROM, or can I flash to either RAM or ROM only,
10:31:08peturI think he refers to run from rom or ram
10:31:11webguest21forgive my ignorance, to me RAM and ROM are two seperate places, and I can either flash to one or the other or both
10:32:22Bagderrwebguest21: you flash to the flash memory
10:32:30Bagderrsomtimes referred to as flash rom
10:33:43peturthe latest unofficial h1x0 bootloader has some (confusing) options to flash rockbox itself iirc
10:33:48webguest21but do I have to flash to both RAM and ROM when flashing firmware
10:33:51gevaertsBagderr: (responding to logs) I still hope for one more application.
10:34:37Bagderrah right, the usb audio?
10:34:50Bagderrwebguest21: just follow the instructions and you'll be fine
10:35:03Bagderryou can't "flash to ram"
10:35:27webguest21ok, but the instructions say i can flash to RAM or ROM
10:35:40BagderrI don't think that's what it says
10:36:04linuxstbgevaerts: Have you heard from the usb audio person recently?
10:36:55webguest21to clarify, I flash to RAM and then to ROM as well
10:37:01gevaertslinuxstb: not since friday (I think). He seems to be online now
10:37:06gevaertsphinze: ping
10:37:20phinzehey hey
10:37:27gevaertsgood morning :)
10:37:55linuxstbwebguest21: It talks about flashing the "ram image" and "rom image" to flash memory. Which is different to "flashing to ram"
10:38:02phinzeindeed... up late pounding away on gsoc apps
10:38:09peturwebguest21: what are you trying to do? just installing rockbox?
10:38:39gevaertsphinze: great :)
10:38:59webguest21ok, so I can flash either the rom or ram image to the same place, right ?
10:39:06phinzegevaerts: actually running behind on my rockbox one ... i'll move my specific questions over to a pm
10:39:15markungevaerts: you need this DFU tool for your M6SL:
10:39:16 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
10:39:21 Quit Chronon ("ZZZZzzzz")
10:39:35peturwebguest21: what are you trying to do? just installing rockbox?
10:39:57gevaertsphinze: sure. You still have more than 15 hours...
10:40:16webguest21petur; yes, I have installed it fine, just trying to clarufy the flash procedure to another noob like myslef
10:41:04peturwebguest21: flashing the ram or rom image is only when you want to put rockbox completely in the flash. Normal install is just flashing the bootloader
10:41:51wpyhHow about the DFU tool on Linux?
10:42:09Bagderrwpyh: that's one of the first things to fix really
10:42:09wpyhI tried it but it could not flash the firmware to the M6
10:42:37peturwebguest21: it is not 'flashing to ram' but 'flashing a ram build' ie a build that runs from ram (<-> a build that runs from rom)
10:43:55webguest21thanks for your help.
10:44:12 Quit webguest21 ("CGI:IRC")
10:44:30wpyhBagderr: I tried to look for a free usb analyzer (someone at Meizume said that you can use a usb analyzer in Windows) but couldn't find any
10:45:26Bagderrand they say DFU is supposed to be a standard ;-)
10:46:05wpyhsomeone at Meizume said that Samsung's DFU is not standard :p
10:48:44wpyhwait, I'll look for that post
10:48:56*gevaerts hopes it is not standard. His first project was going to be to work out the differences
10:49:24Bagderrwell, hopefully it just breaks it at some small point
10:50:13Bagderriirc, markun said the SoC boots in big-endian mode and I wonder if the DFU then needs to speak that
10:50:33 Quit cool_walking_ (Remote closed the connection)
10:50:33Bagderrthe openmoko dfu page says it doesn't work in big endian
10:51:15markunBagderr: according to kgb2008 it's very different from the dfu standard. We'll see after gevaerts analyzes the USB trafic during a firmware flash.
10:51:35Bagderrso the "dfu utils" name is just misleading then
10:51:54Nico_PJdGordon: you're having issues with playback?
10:52:24Bagderrkgb2008 is a man with many facts so I'm sure that's correct
10:52:25amiconnmarkun: There was some talk regarding the old rockchips in here several weeks ago. There was also a link to a datasheet
10:53:26markunamiconn: I only have the datasheet for the RK26xx
10:53:33 Join Spex [0] (
10:56:02 Join daurnimator [0] (i=daurn@unaffiliated/daurnimator)
10:57:24wpyhmarkun: you have the datasheet for RK26xx?
10:57:25Nico_PBagderr: about GSoC apps, knowing how students tend to do things at the last minute, I expect a couple new ones today (not counting phinze's)
10:57:49*phinze isn't counted :(
10:57:52Bagderryeah, I guess we'll find out!
10:58:18*gevaerts thinks that Nico_P means 'unexpected applications' :)
10:58:28Nico_Pexactly :)
10:59:29*phinze knows this, but always jumps at the chance to be meek and lowly ;)
10:59:30Nico_Pamiconn: don't know if you've seen: I've found the cause for the end of playlist bug but I need to test the fix
11:00:48Nico_Pamiconn: if you're interested in helping test: comment out playback.c:1947 and report any unexpected stopping (I think one might happen on codec change towards the end of the playlist)
11:01:40wpyhmarkun: Thanks
11:01:52 Join Jeton [0] (n=4f7ebe95@gateway/web/cgi-irc/
11:02:08Nico_Pbtw, I'm the first one to agree it sucks that the playback engine still can't be called stable :(
11:02:35 Quit Vampiryan ()
11:03:18JetonWith the commit of FS #8308 do themes need to change some parametres in order to resize AlbumArt? I've got the latest build and I don't see any resizing. Or i'm missing something?
11:03:33Nico_PJeton: it's only available to plugins for now
11:03:58Jetonjust sliding puzzle right?
11:05:18Jetonallright. Any idea when will it apply for AA? Just curious.
11:05:29Nico_Pnot really, no
11:05:31 Join tvelocity [0] (
11:05:43Jetonokay, thanks .
11:05:49Jetonokay, thanks .
11:05:56Nico_Pit'll be when someone decides to write a version we can use in the core
11:08:23 Quit Jeton ("CGI:IRC (EOF)")
11:08:55amiconnmarkun: I confused rockchip with another DAP SoC: the ATJ2085 (check logs from 8 August 2007)
11:09:24wpyhATJ is Actions
11:09:33wpyhAlso a China-based manufacturer
11:12:33phinzeso, lets say i managed to get USB Audio implemented such that your PC sees your DAP as a USB Soundcard and can send its audio output to it
11:13:09phinzewould it be Absurdly Difficult to mix the bleeps and bloops coming out of your daily PC activities into the music audio output in rockbox
11:14:11Bagderrmy daily PC has no "bleeps and bloops" ;-)
11:14:20Bagderrisn't that mostly a quetion of what your OS supprots?
11:14:25phinzebecause, you see, if one could mix the two, this could be a fairly useful use case... one pair of headphones for your music and your IRC pings
11:14:27***Saving seen data "./dancer.seen"
11:14:30*Bagderr is a typing master
11:14:39*gevaerts translates:
11:14:58gevaertsBasically how hard is it to implement mixing of two audio streams within rockbox ?
11:15:12wpyhphinze: that's a nice feature :)
11:15:19Bagderrah within Rockbox? I thought the mixing would be done in the OS side of things
11:15:20gevaertsphinze: always put this in terms that rockbox people understand :)
11:15:42*phinze took the long way around
11:15:55Bagderrbut now I see, Rockbox would run as normal and just get the "PC beeps" passed in?
11:16:08phinzeBagderr: right because the OS would be sending all its audio output to what it sees as a USB soundcard
11:16:14phinzewhich would actually be your DAP running rockbox
11:16:34Bagderrso when you run winamp on the PC to play music, you'll get to enjoy twice the music!
11:16:43phinzeand rockbox would have to mix the USB Audio input stream along with its normal musical output
11:17:03phinzeBagderr: that's only for power users ;)
11:17:25 Join webguest82 [0] (n=d9e1ca0d@gateway/web/cgi-irc/
11:17:27BagderrI figure the sound comes in some kind of pcm stream?
11:17:46webguest82Hi, once i have flashed my player, can I delete the .rockbox folder ?
11:18:04Bagderrwebguest82: no
11:18:24 Quit webguest82 (Client Quit)
11:19:03gevaerts... Possibly forced to one specific sample rate/pcm format/sample size setting (so the host OS would do the needed resampling)
11:19:43gevaertsSo you can assume whatever pcm format you like
11:19:46Bagderrso it's basically a constant cross-fade
11:20:10Bagderrthe feature I mean
11:20:37*gevaerts doesn't know too much about this side of things, but thins that Bagderr is correct
11:20:42phinzeBagderr: more or less, except i would imagine less decoding needs to be done for the PC Audio Stream than for the MP3
11:21:06Bagderrright, it would bascially be to mix in another pcm stream with the decoded song
11:21:38*gevaerts nods
11:21:38phinzeyep. so... does that fall in the realm of the vaguely possible?
11:21:44peturjust like voice...
11:21:52Bagderrit certainly sounds possible to me
11:21:57phinzeooo, that's what rockbox already does with voice?
11:22:07phinzemixes it in with your music stream?
11:22:45phinzenice. that's promising.
11:23:55Bagderroh a new application
11:24:11Bagderrand gone...
11:24:32*gevaerts was slightly too late :)
11:24:55Bagderrwas just one of those silly nonsensical ones
11:25:20wpyhWell, doesn't rockbox need to process the USB data?
11:26:32gevaertswpyh: rockbox can tell the host what exact format it wants, so it can ask for something that doesn't need further processing
11:26:33*phinze has been up much too long to catch these sorts of things :)
11:26:53*gevaerts guesses. Could phinze be a student ? ;)
11:27:15phinzeworking at 4:30am the night before a deadline? sounds about right.
11:27:46JdGordonNico_P: hey, yeah playback went toatly fubar this arvo.. but im not using current svn so havnt done any proper reports
11:28:07phinzewpyh still has a point, though... if the processing of the isochronous (sp?) transfer takes too much overhead that might kill the entire idea
11:28:23JdGordonI'm also thinking it may be a dead file system because the wps kept saying it was the same track and the progress was ok, and eveentually it got back to the playing track
11:28:44phinzeand may even kill your regularly scheduled playing music
11:29:12phinzethoughts, gevaerts?
11:29:38gevaertsTrue, but I don't expect it will on 'normal' codecs. There is still a lot of cpu available
11:29:42Nico_PJdGordon: ok. try to at least use something after r16955 ;)
11:30:10JdGordon16891 is the version on my sansa..
11:30:16*JdGordon needs to update :p
11:30:23gevaertsAnd we're not too worried about power usage here. We have 500mA if we want them :)
11:30:27phinzegevaerts: but aren't you already seeing the USB stack bugging out when the CPU is busy handling other things?
11:30:33phinzemmm... power
11:30:59gevaertsphinze: yes, but that's a bug ;) And it only happens for transfers going the other way
11:31:07phinze...unlimited POWER!
11:31:11phinzeerr... sorry.
11:31:42phinzealright, i'll take your word for it that there's CPU to spare
11:33:09 Join ap0 [0] (
11:42:20 Part Spex
11:44:12Bagderr... gsoc mentor introductions couldn't be much more uninteresting to post to the mentors list
11:44:39gevaertsAnd they still go on...
11:44:41 Join Spex [0] (
11:44:42*petur labeled them as spam...
11:44:54Bagderrwell, 130+ orgs with 10+ mentors in each... they can go on forever!
11:45:09peturmaybe we should start mailing too?
11:45:43phinze"Hello, my name is X and I would like to apply for GSoC 2008! Now, what do I do?"
11:45:55*gevaerts waits for the time when petur's mail system classifies all petur's mail as spam :)
11:46:14peturpeace, finally ;)
11:47:50Bagderrphinze: the list is only for mentors, so they don't apply ;-)
11:48:15Bagderr"helly my name is N I'm mentor in project P"
11:48:32 Quit ap0 ("Bye.")
11:48:36phinzeahh well so you avoid those at least... the Gallery project set up a GSoC specific mailing list with mentors and students
11:48:40Bagderras if anyone outside that project cares...
11:48:54Nico_PBagderr: don't worry, the student ML is much the same
11:49:04phinzedozens. dozens. of nearly identical emails.
11:49:06 Part Spex
11:49:36 Join Spex [0] (
11:49:38phinzeat least in my svn-announce list the revision goes up every time!
11:49:43gevaertsI count 401 unique senders now
11:49:47Nico_Pdoes google tell everyone to introduce themselves?
11:50:08BagderrNico_P: no, but they didn't shoot down the idea when someone suggested it
11:50:45gevaertsThey probably missed that mail, and when they noticed there were already 100 of them
11:50:57peturwhich is stupid, given that most of those mails will be stored on gmail...
11:51:03Bagderrno, they said "it might be a good idea" or similar
11:51:45Bagderroh well, it's fairly easy to just not read them
11:51:59*petur nods
11:52:26phinzeplus, google does have More Storage Than Humans Can Comprehend
11:53:57phinzewho needs efficiency when you can throw a 500 node cluster around?
11:55:40*phinze wonders if the enterprise folks make fun of the embedded folks as much as the other way around
11:56:16Bagderrof course not, we are the cool guys and they know it!
11:56:36 Join DavidS1 [0] (
11:57:24phinzethough i suppose they would just invoke the "whose is bigger" type banter
11:57:47JdGordonlogbot: !seen barrywardell
11:57:58*JdGordon thinkis he got that wrong
11:59:07GodEater!seen barrywardell
11:59:21GodEatereither that or logbot only responds to certain people ? :)
11:59:33amiconnJdGordon: /msg logbot seen <nick>
11:59:48***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
11:59:58JdGordonor logbot isnt registered still?
12:00:23JdGordonyep :(
12:00:44JdGordonok.. anyone seen barry lately?
12:00:45amiconnSet unfiltered, then you'll get the answer
12:01:31JdGordonhas anyone else with a PP target had problems loading the OF after buschels power saving commit?
12:01:47JdGordonseems there is a problem with DEV_USB0 at least on e200
12:02:00*amiconn certainly won't notice
12:02:13JdGordoncommenting out its disable line fixes the problem.. I want to commen it out in svn...
12:02:32 Quit Nevtus ("Gone")
12:02:33JdGordonamiconn: you never use of for usb?
12:02:51amiconnno, why?
12:03:11amiconnIpods have emergency diskmode, and the H10 has in-rom diskmode too that comes before the rb bl
12:03:11*linuxstb suspects JdGordon is talking Sansa, and amiconn isn't...
12:03:29JdGordonsupposedly this problem happend on both an ipod and a h10
12:03:42LinusNamiconn: wouldn't that count as "of"?
12:04:14BagderrI count everything outside of Rockbox as OF
12:04:31JdGordonFS #8506 is the report
12:05:02JdGordonsorry, 4g and mini supposedly had the issue
12:05:12LinusNstill, in this case, there is a difference between "of" and "of"
12:05:41 Join Nevtus [0] (n=Nevtus@unaffiliated/nevtus)
12:05:42amiconnJdGordon: Besides, on mini I mostly use rockbox usb nowadays
12:06:06 Part Spex
12:06:11 Quit kugel (Remote closed the connection)
12:12:32 Quit DavidSG (Connection timed out)
12:15:10JdGordondoes anyone mind me adding usb_enable(true); in the usb thread before it tries to reboot when not using rocbox usb? (assuming this fixes the problem)?
12:15:49soapJdGordon, I have no trouble booting into OF on my iPod 5th gen.
12:16:52gevaertsJdGordon: the problem with that is that you don't want the usb_core_init() from usb_enable
12:17:15JdGordonwill that cause problems if its going to reboot?
12:17:32linuxstbLinusN: Do you have a backup of any of the Rockbox_LLorean theme files from the forums? IIUC, Llorean has a copy, but on the hard disk of a broken laptop...
12:20:45 Join kugel [0] (i=kugel@unaffiliated/kugel)
12:21:18JdGordongevaerts: so is rockbox usb safe to use on the e200?
12:21:32*JdGordon thinks it might be easier to use that than try to fix the bootloader :p
12:22:04gevaertsJdGordon: there is still FS #8663 unfortunately
12:22:42kugelJdGordon: I hat this issue yesterday too (FS8506)
12:22:45linuxstbIsn't full-speed reliable though?
12:23:05JdGordonkugel: on which target?
12:23:22gevaertsAt least desowin saw the problem once on full speed on e200...
12:23:33kugelWith a current build
12:24:01 Join roolku [0] (
12:24:11kugelIt worked the previous times though, yesterday was the first time I experienced that crash
12:28:44JdGordongevaerts: in usb_init() staright after usb_init_device() there is usb_enable(false); shouldnt it be checking if usb is connected before disabling it?
12:31:28gevaertsGood question. For PP that call is not needed there (since it basically does the same as the end bit of usb_init_device(), but I'm not sure about targets with USB-ATA bridges
12:33:15 Join Thundercloud [0] (
12:44:02JdGordonarg... I was accidently copying the wrong file across and getting pissed off when nothing was changing :'(
12:44:19JdGordonseems I've got the bootloader playing again... need to figure out what is needed and what isnt now
12:44:38*gevaerts usually edits in one tree, compiles in another, and copies files from a third
12:45:30phinzethe Forest Method
12:45:52JdGordoncomputers should be smart enough by now to know what we mean!
12:46:54JdGordongevaerts: why is dmesg showing a rockbox usb device when rockbox usb isnt enabled?
12:48:28gevaertsIt has to show something to be able to get any current from the bus. I agree that it isn't really needed just before rebooting, but it's kind of tricky to avoid this
12:49:10gevaertsIt only has to reboot if the host does usb (i.e. not when connecting to a wall charger), so you need basic enumeration support
12:49:53JdGordonany issues commiting ?
12:50:59JdGordonamiconn: LinusN: ^ will removing usb_enab;e(false) from the bootloader cause any issues on non PP targets?
12:51:26 Nick fxb__ is now known as fxb (
12:52:46gevaertsJdGordon: I see no issues, but I can't do any testing or in-depth review now. I guess we'll notice if it makes things break ;)
12:53:13JdGordonexcept, it will only affect bootloader, so any issues might not show up for months :(
12:55:05gevaertsI can test on my players tonight (gigabeat, ipod 5G and c200, I' rather not touch h300)
12:56:46JdGordongetting a nice collection there :)
12:56:59JdGordonspeaking of which.. my mr500 battery finnaly went dead :D
12:57:12gevaertsI'll test on the meizu once the port is done :)
12:57:18*phinze wanders into the kitchen for a snack, figuring if depriving body of sleep best to at least give it food
12:57:38*gevaerts goes to have lunch as well
12:58:36*JdGordon notices the time and doesnt do anything about that commit for a few hours to hopefully get a respons
13:10:28 Join n1s [0] (
13:14:30***Saving seen data "./dancer.seen"
13:21:19 Join dabujo [0] (
13:21:24 Join ali_as [0] (
13:25:36 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
13:33:06 Join krz [0] (
13:35:50pondlifeNico_P: Have you written down the correct sequence of events for a track transition?
13:36:03Nico_PI have it on a sheet of paper
13:36:15pondlifeAny chance of it making the wiki?
13:36:42pondlifeLooks like playlist_next needs a partner to do a non-advancing get?
13:37:01*pondlife suspects playlist_next() should be renamed playlist_advance()....
13:37:59pondlifeI'd very much like to get a proper sync on the playlist index, I'm sure you would too...
13:38:42Nico_PI also had to reopen a bug when committing r16955
13:38:48pondlifeWhich one?
13:38:55n1sNico_P: does FS #7033 make sense anymore with viewports now supported in wps?
13:39:29Nico_Ppondlife: FS #8206
13:39:34linuxstbn1s: IMO it could still be useful.
13:39:36Nico_Pit was fixed by r16019
13:39:51pondlifeNico_P: Yep - that's a playlist index confusion one too
13:40:28Nico_Pn1s: I think it could still be useful but it needs an improvement with scrolling lines
13:40:52n1slinuxstb, Nico_P: ah, for the same wps but using different fonts :)
13:40:55Nico_Ppondlife: I'll try to make a diagram and put it on the wiki, yes
13:41:07pondlifeDo we have a conflict of some kind in there?
13:41:23pondlifei.e. a must happen before b, b must happen before c, c must happen before a
13:41:39pondlifewhere playlist_next() is one of them :)
13:41:56Nico_Pcurrently, yes. but I think it can be solved
13:42:02linuxstbn1s: Yes, or simply when you want a different line spacing...
13:42:25 Join MethoS- [0] (
13:42:36Nico_Pthe problem I fixed in r16955 was that audio_rebuffer() was called by audio_checj_new_track before playlist_next() had been called
13:42:47pondlifeYes, I saw your notes
13:43:03linuxstbn1s: But i would only want it if the code impact is minimal (which I think it was before viewports) - if it complicates things, I would say to forget it.
13:43:04pondlifeBut then tried to work out the proper sequence in my head and fell over
13:43:14Nico_Pit could have been fixed with a correct value for last_peek_offset but in the current state of things, it would have added even more special casing
13:43:24n1slinuxstb: yes, that too, btw have any of you wps hackers seen fs#8517?
13:43:30pondlifeBetter to fix it properly.
13:45:07Nico_Ppondlife: I'd like to rewrite audio_check_new_track to better separate distinct cases... currently they are all intertwinded
13:45:26pondlifeI'd like to see audio_check_new_track() die :)
13:45:32pondlifeslowly and horribly
13:46:03pondlifeThat's exactly the sort of routine that gives poor playback.c a bad name
13:47:41Nico_Pindeed. I think it's really the weak point currently
13:48:14Nico_Pthe weakest link, if I dare say :)
13:49:15 Quit TaylorKillian (Read error: 110 (Connection timed out))
13:49:20pondlifeYou might want to not force the playlist index to sync with the rest of the WPS...
13:49:56pondlifeIt's useful to be able to skip forward repeatedly and see the playlist index(and maybe filename) increment quickly, but not have to wait for metadata to buffer
13:50:24pondlifeProbably not a playback/playlist thing, but something useful that got a bit lost in MoB
13:50:25Nico_PI agree. that's why I had put playlist_next in audio_skip
13:51:08pondlifeAlso, if we could add how MASCODEC does transitions to your diagram, that would be useful.
13:51:35pondlifeWhen I say, we... I really mean someone with a MAS
13:51:39Nico_PI had a quick look and the lack of a codec seems to make things much simpler
13:51:49 Join ap0 [0] (
13:53:02linuxstbJdGordon: Did you install the userfs tgz file? My guess is that needs to go to the root as well, and the author forgot that from the "full install" archive.
13:53:13linuxstbI mean usrfs-0.1.tar.gz
13:53:14Nico_Pmaybe to get the playlist index in sync with the WPS we should make the WPS not read the index directly but instead use the ID3 struct
13:53:28JdGordonlinuxstb: there is a 20mb userfs file which is in the root
13:53:51Nico_Pthere's an index field there that's always zero on SWCODEC and that is used for something I don't really understand on HWCODEC
13:54:03linuxstbJdGordon: So there is... I missed that.
13:54:26JdGordonmy desktop has no problems mounting it
13:55:21Nico_Ppondlife: do you need a quick briefing on the track skipping procedure or was it mainly for documentation purposes?
13:55:31pondlifeBit of both really
13:55:47pondlifeNo big rush, I won't have time to play today
13:56:18pondlifeBut if such a document had been in our possession beforehand, we'd have wasted less time...
13:56:19Nico_Pfne by me, I also have some school work to do :)
13:56:36pondlifeOK, just wondering if I needed to work it out from scratch or not.
13:56:55pondlife"Not" being a good answer. :)
13:57:26Nico_Pin short, it's always initiated by the codec, through codec_request_next_track_callback
13:57:48Nico_Pthat leads to sending (not posting) Q_AUDIO_CHECK_NEW_TRACK to the audio queue
13:58:10Nico_Paudio_check_new_track then gets called and replys
13:59:01Nico_Pa manual skip is triggered by forcing ci.new_track to 1, which will make the codec request the next track
14:00:17Nico_Ponce the skip is finished, codec_track_skip_done is called, which posts Q_AUDIO_TRACK_CHANGED to the audio queue, at which point audio_finalize_track_change is called
14:01:13 Join DerDome [0] (
14:02:57Nico_Ppondlife: does that help?
14:03:27pondlifeRemind me about post vs send...?
14:03:33pondlifei.e. which blocks?
14:03:49 Part LinusN
14:04:18Nico_Psend blocks the caller until it gets a reply
14:04:27Nico_Pit's the synchronous one
14:04:49 Quit mrfree ("Leaving")
14:04:58Nico_Pso audio_check_new_track better be as fast as possible
14:05:05pondlifeThe codec calls codec_track_skip_done when it has decoded the last of the data?
14:05:41 Quit GodEater (" ajax IRC Client")
14:05:46pondlifeAnd... manual skips aren't really relevant , they just trigger the codec to kick off the whole thing?
14:05:52 Join GodEater [0] (i=c2cbc962@rockbox/staff/GodEater)
14:06:56Nico_Pcodec_track_skip_done is called as soon as audio_check_new_track replies to the codec
14:07:08pondlifeSo my old terminology about "codec transition" and "WPS transition" still makes sense.
14:07:25Nico_PI call it the transition window :)
14:07:53Nico_PI simplified the end of the skipping a bit
14:08:08 Join MethoS-- [0] (
14:08:20Nico_Pwhen the skip is manual, Q_AUDIO_TRACK_CHANGED is sent immediately
14:09:12Nico_Pwhen it's automatic, that message is delayed until PCM has effectively changed tracks
14:09:13pondlifeSay the codec_request_next_track_callback occurs and audio_check_new_track returns, but there's still 8 seconds' of audio to play out before the WPS changes. What's being executed before the codec_track_skip_done call?
14:09:21Nico_Pcodec_track_skip_done is what handles crossfading and gapless
14:10:25pondlifeI think a full timeline with all routines and Q_... stuff would be most useful
14:10:25Nico_Pcodec_track_skip_done leads to enabling the pcm position callback, which will update the wps progressbar
14:10:55pondlifeAh, so codec_track_skip_done is still part of the "codec transition" and before the "WPS transition"
14:10:58Nico_Pand it also enabled a pcm track changed callback, which is what will post the Q_AUDIO_TRACK_CHANGED message to the audio queue
14:11:20pondlifeGotcha, Q_AUDIO_TRACK_CHANGED is the point the user sees a change
14:11:34pondlifeAnd the point when the playlist index should really incremenet
14:11:39pondlife(or however you spell it)
14:12:00pondlifeI mean, to the user interface (including resume point)
14:12:00Nico_Pit should, but making it increment in audio_check_new_track is simpler in most cases
14:12:44Nico_Pactually the only case where you want to advance the playlist in audio_finalize_track_change is on an automatic skip
14:13:11Nico_Pin all (I think) other cases, you need to call playlist_next in audio_check_new_track or before
14:14:07pondlifeI think you need to have a playlist API change so the playlist index is more controlled..
14:14:35pondlifeThe outside world should see the increment only when the Q_AUDIO_TRACK_CHANGED occurs
14:14:50pondlife(which is early on manual skip, anyway)
14:14:57Nico_PI agree
14:15:13pondlifeProblem is the way playlist_next() works... and the rest of the complicated playlist API.
14:16:12pondlifeI may have already said, but I wonder if some of the playlist API is only used for SWCODEC, and could be scrapped...?
14:16:33Nico_PI think wa can get away with it by carefully rewriting audio_check_new_track
14:17:25Nico_Pmaybe simply with a skip_type enum and a switch construct on it
14:18:19Nico_PI think I have a rather good idea of what needs to happen in each case. the problem is they are completely mixed together ATM
14:21:43 Quit HEx (Remote closed the connection)
14:21:59 Quit perrikwp (" ajax IRC Client")
14:23:58*Nico_P would like to come up with some sort of regression testing codec
14:24:20Nico_Psomething like test_codec that tests playback behaviour instead of testing performance
14:24:57Nico_Phmm actually it would be the opposite of what test_codec is :)
14:26:00 Join miepchen^schlaf [0] (
14:26:26 Quit miepchen^schlaf (Client Quit)
14:26:40 Join miepchen^schlaf [0] (
14:27:38 Quit Febs (Read error: 110 (Connection timed out))
14:28:10amiconnNico_P: Looks like id3.index is set to the actual playlist index on hwcodec (and the comment in id3.h says so too)
14:28:12 Quit MethoS- (Read error: 110 (Connection timed out))
14:30:01Nico_Pamiconn: what confused me is track->id3.index = steps; in new_file()
14:31:38Nico_Pbut I think the WPS should be made to use that value instead of directly using playlist_get_display_index()
14:31:52*phinze is finished with his GSoC app... now to get it to look decent in those damn plaintext <textarea>s
14:32:17linuxstbuuencode application.doc
14:32:22Nico_Pamiconn: would that be correctly simulated in an HWCODEC sim?
14:32:27*linuxstb hides
14:32:40amiconnMost probably not
14:34:52JdGordonamiconn: did you see my pastebin before? i want to double check that change wont affect non PP targets before commitnig
14:35:23Nico_Pamiconn: then I'd need some help in testing
14:35:32Nico_PI'll try though
14:36:53 Part Bagderr
14:37:13 Join barrywardell [0] (n=barrywar@rockbox/developer/barrywardell)
14:37:26 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
14:37:36 Quit linuxstb (Nick collision from services.)
14:37:46JdGordonhey barrywardell
14:38:20 Join bughunter2 [0] (
14:38:52barrywardellhi JdGordon
14:39:22JdGordonare you still having problems loading the OF on sansa?
14:39:52JdGordonsvn bootloader looks like the pin detect is broken... i want to commit which hopefully fixes it?
14:41:59*phinze winces, hits submit
14:43:35 Quit phinze ()
14:45:04*n1s want's to make a joke about garbage collection but supresses the urge...
14:45:11Nico_Phmm, id3.index is zero-based on HWCODEC
14:45:20n1snot having seen any applications of course
14:45:51 Quit MethoS-- (Read error: 113 (No route to host))
14:48:25 Join nplus [0] (
14:48:34barrywardellJdGordon: I thought I fixed the of loading
14:48:56JdGordonnope.. was still broken on mine
14:48:59 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
14:49:07JdGordonendless looping becasue usb detect seemed to be broken
14:49:36barrywardellah, my problem was different
14:49:43JdGordonand FS #8506
14:49:55barrywardellit wasn't able to load the of at all in my case
14:51:54barrywardellnow that you mention it, I think the pin detect might not have worked for me
14:51:55JdGordonthat patch fixes it.. but I have no idea if it will break non-pp targets
14:54:17*JdGordon rekons it probably is safe...
14:54:26 Join ChristopherW [0] (
14:56:34barrywardellit looks reasonable from a pp point of view anyway
14:56:52barrywardellI think you should commit provided it doesn't break other devices
15:02:41barrywardellJdGordon: I can confirm that bootloader usb detection doesn't work on my sansa with current svn. testing now to see if your patch fixes it
15:02:50 Quit nplus (Remote closed the connection)
15:03:09JdGordonhopefully it does.. something fell on the enter key while i had "svn ci" on the console :p
15:03:18JdGordoni mean.. :D
15:04:58 Join nplus [0] (
15:06:32 Join FOAD_ [0] (
15:07:53*JdGordon thinks we should probably release a new bootloader for e200's after that commit
15:08:16 Join codin [0] (
15:08:50gevaertsI think we'll need a #rockbox-mentors channel soon. While it's all very well to suggest application improvements in public, at some point we will have to state negative things as well, and I think a private channel is better for that
15:09:51barrywardellJdGordon: yep. it fixes things for my sansa too. Also, I agree that it most likely shouldn't do any harm to other devices
15:09:54JdGordonnot if you tell everyone about the private channel :p
15:10:00codindo you know how was the manual pdf file created and built ? - what tools were used ?
15:10:04 Quit Christopher1 (Read error: 110 (Connection timed out))
15:10:13barrywardellJdGordon: why release a new bootloader?
15:10:43JdGordonbecause the old ones have an issue when rockbox disables that DEV_USB0
15:10:50gevaertscodin: it uses LaTeX tools, but I don't know the exact requirements
15:11:23codinI saw that all the fonts are T1, and are embedded in the document
15:11:29barrywardellJdGordon: as far as I know, the current one on the download servers was from before all that stuff was added
15:11:43codinI'm interested in some guidelines on how to do that
15:11:55codinor a template for the document, if possible
15:11:57JdGordonbarrywardell: so its enabled and not disabled?
15:12:50barrywardellJdGordon: I think the one on the download servers was before we did anything with DEV_USB0, which probably means it is left enabled
15:13:47codingevaerts, do you know a way I could find out ? somebody to contact ? -thank you
15:13:53gevaertscodin: this isn't really on topic here, but if you're looking for something like a word template, this is _very_ different
15:14:03 Quit EspeonEefi ("さよなら")
15:14:04 Join mf0102 [0] (n=michi@
15:14:13 Join ByeongKeon [0] (n=skysign@
15:14:30codingevaerts, _not_ word !, I am looking for a latex cls or sty file/template
15:14:33***Saving seen data "./dancer.seen"
15:14:36 Join LambdaCalculus37 [0] (n=44a0430f@gateway/web/cgi-irc/
15:14:43codinor some hints/pointers
15:14:48 Quit FOAD (Read error: 110 (Connection timed out))
15:14:49 Nick FOAD_ is now known as FOAD (
15:15:21 Join perrikwp [0] (i=9821663a@gateway/web/ajax/
15:15:42gevaertscodin: ok :) If you download the rockbox source from, it should contain the manual
15:16:08codinthank you very much, I'll have a look
15:17:52barrywardellJdGordon: just tested to be sure and the current bootloader on works just fine, usb detection and all :D
15:18:33 Join Arathis [0] (
15:19:03 Quit barrywardell ()
15:22:29linuxstbcodin: This page may help you as well -
15:22:59 Join jgarvey [0] (
15:26:07codinlinuxstb, thank you very much
15:27:17gevaertsByeongKeon: In your application you mention that you developed a Linux USB client driver for a MP3 Player DIY. Are there online references somewhere that we could look at ?
15:28:06ByeongKeongevaerts: hello gevaerts.
15:29:17ByeongKeongevaerts: here is
15:30:05ByeongKeongevaerts : you can download source code from
15:31:39ByeongKeonit was written in korean, if you want to read that, I'll translate it.
15:32:01gevaertsNo problem. C is still C ...
15:32:32 Quit Rob2223 (Read error: 110 (Connection timed out))
15:33:23ByeongKeongevaerts, I have a question.
15:33:34gevaertsSure, go ahead
15:34:13 Join DerPapst [0] (
15:35:03ByeongKeononly, you ask me about my application and leave comments.
15:36:13 Join Rob2222 [0] (
15:36:21ByeongKeonit looks, other mentors are not interesting my applicattion.
15:37:34gevaertsThey are, but for some reason I seem to be asking the questions first. The same is true for most other applications. I expect this to change next week when the evaluation period starts
15:37:49*gevaerts has a meeting now
15:41:47Nico_PByeongKeon: we're all interested in most of the applications, but gevaerts has been doing most of the work
15:41:56Nico_Pit doesn't mean we're not following ;)
15:42:20ByeongKeoncan I ask one more?
15:42:27*petur opens one eye
15:42:42Nico_PByeongKeon: of course
15:42:43JdGordonmove along.. nothing to see here...
15:43:22Nico_Pamiconn: I hop you'll like r17018 :)
15:43:39 Quit Arathis ("Bye, bye")
15:44:53ByeongKeonI'd like to talk other student who submit application, but, I'm only a student who submit application in here.
15:45:39Nico_Pthey come by too but don't usually stay very long
15:46:02peturByeongKeon: did you try #gsoc ?
15:46:54ByeongKeonhello petur, #gsoc channel is for all student.
15:47:20ByeongKeonI mean a student who submit to Rockbox.
15:49:07Nico_PByeongKeon: your best chance to get to talk to them is to stick around and wait for some gsoc talk
15:49:20 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
15:49:26 Join mcflow [0] (
15:49:36 Quit linuxstb (Nick collision from services.)
15:49:38 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
15:50:52ByeongKeonNico_P: Tomorrow, I'll do as you told.
15:52:20ByeongKeondid you see domonok1 ? he told me yesterday, he will leave some comments on my application.
15:53:01ByeongKeonSo, I'm here, to answer that. but, he didn't leave anything.
15:53:17Nico_Phe's around quite often
15:54:06 Quit codin (Read error: 104 (Connection reset by peer))
15:54:20 Join codin [0] (
15:54:20 Quit codin (Read error: 104 (Connection reset by peer))
15:54:33 Join codin [0] (
15:54:41ByeongKeonI'm going to wait for him.
15:57:47DavidS1ByeongKeon: I am also a student that applied for a gsoc with rockbox
15:57:57 Join Febs [0] (n=chatzill@
15:58:22ByeongKeonyou are. hello, DavidS1.
16:00:06*DerPapst waves to the other gsoc appliants :-)
16:00:39DavidS1DerPapst: did you get your ppm veiwer done after?
16:00:56Febsgevaerts: I'm not sure if the information in this forum thread is of any use to you, but I thought I'd bring it to your attention:
16:00:58DerPapstlast compile running right now i hope.
16:01:17*DavidS1 wishes luck all around
16:03:00 Nick DavidS1 is now known as DavidSG (
16:03:31DerPapstmeh... native gcc complains about something what arm-elf-gcc is fine with ^^
16:08:28linuxstbDerPapst: Different versions of gcc often give different warnings...
16:08:50linuxstbe.g. 4.0 vs 4.2
16:15:39DerPapstfixed :-)
16:16:27 Join FOAD_ [0] (
16:20:44 Join cameron12354 [0] (
16:22:27cameron12354hey everyone
16:22:45cameron12354can i get any else for rockbox here?
16:23:01cameron12354like games programs and what not
16:23:03 Part cameron12354
16:25:34 Join J-23 [0] (
16:29:11 Quit FOAD (Read error: 110 (Connection timed out))
16:29:11 Nick FOAD_ is now known as FOAD (
16:30:08 Join perrikwp|class [0] (i=98213489@gateway/web/ajax/
16:33:31Nico_Ptrack skipping is fscked up :/
16:35:33 Quit perrikwp|class (" ajax IRC Client")
16:44:46 Join disorganizer [0] (
16:46:11 Quit Llorean ("Leaving.")
16:48:16 Quit ender` (Read error: 104 (Connection reset by peer))
16:48:23 Join Lynx_ [0] (
16:48:27 Join ender` [0] (
16:48:38 Join mcuelenaere [0] (
16:49:01LambdaCalculus37Nico_P: Skipping back through tracks is especially forked up.
16:49:15LambdaCalculus37Sometimes it still skips *forward*.
16:49:28Nico_PLambdaCalculus37: yes, I've just seen how bad it is
16:49:48 Join Piero_Scarufii [0] (
16:50:08Piero_ScarufiiDoes Rockbox really not support PNG or JPEG album art?
16:50:16Piero_ScarufiiI'm just wondering if CPU power is really such an issue that they would do that.
16:50:22disorganizerregarding FS# 7033: i think this still makes sense at least until we get a "vertical centering" of a text line inside a viewport :-)
16:50:36LambdaCalculus37Rockbox only supports .bmp artwork.
16:50:45Nico_PPiero_Scarufii: we only support bitmap, but we'd like to add support for other formats
16:50:52Nico_Pit's just that noone has done it yet
16:50:52*pondlife quietly wants bmp resize in the core...
16:50:57dionoeaPiero_Scarufii: it's more of a memory management issue AFAIK
16:51:10Nico_Ppondlife: seen my commit?
16:51:16pondlifeThe latest?
16:51:21Piero_Scarufiiok ty
16:51:24Nico_Pyes, the playlist index one
16:51:24LambdaCalculus37pondlife: I wouldn't mind seeing bitmap resize, either.
16:51:33pondlifeIs that what broke track skip?
16:51:48pondlifeI've not got any DAP with me at the moment
16:51:51Nico_Ppondlife: no, it didn't change anything besides the WPS
16:51:55pondlifeAh, ok
16:52:06Nico_Pit's purely the view in this case
16:52:15pondlifeI'm very out of date here...:)
16:53:04Piero_Scarufiidoes rockbox support creative players yet?
16:53:13GodEaternot yet
16:53:47Piero_Scarufiiwhy not... too hard?
16:53:52linuxstbPiero_Scarufii: And "creative players" means lots of different hardware - if/when support is added, it will be for specific players.
16:55:13linuxstbPiero_Scarufii: Rockbox is a community of developers - people join the project and port Rockbox to players they own and are interested in. Only recently have users with the skills started to hack any Creative players.
16:55:57Piero_Scarufiithats fine i use my creative as storage anyways
16:56:03Piero_Scarufiijust wondering
16:56:14Piero_Scarufiii asked the same question here 3 years ago :)
16:58:02scorche|shPiero_Scarufii: for future reference, all the devices rockbox runs on are mentioned on the front page
16:59:24Piero_Scarufiii never visit sites
17:00:11Piero_Scarufiibut thanks
17:00:49 Quit DerPapst (Read error: 110 (Connection timed out))
17:00:57 Join toffe82 [0] (
17:06:16amiconnNico_P: There might be a problem with that commit on hwcodec. I'll check tonight
17:06:17 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
17:07:43Nico_Pamiconn: what kind of problem?
17:08:18Nico_PPiero_Scarufii: there is ZVM port that has made nice progress
17:08:38scorche|shPiero_Scarufii: *never* visit sites?
17:08:49Nico_Pand it has opened the way for ports to other creative players
17:09:02amiconnThere might be wrong indices displayed when playing a dir shuffled (according to wps.c / wps-display.c history)
17:09:45Piero_Scarufiiscorche thats right
17:09:52Piero_Scarufiii never visit sites
17:10:06LambdaCalculus37Piero_Scaufii: Not for anything, but start making it a habit of doing so.
17:12:01 Join mcuelenaere [0] (
17:12:30 Join DerPapst [0] (
17:14:26Nico_Pamiconn: I'll try in the sim
17:14:37***Saving seen data "./dancer.seen"
17:16:49 Quit corevette (Remote closed the connection)
17:16:49Nico_Pamiconn: it seems ok here, but please test it :)
17:18:30disorganizerkugel: your patch breaks massa's patch or vice versa :-)
17:18:49kugeldisorganizer: I know
17:18:56 Quit ap0 ("Bye.")
17:19:38disorganizerkugel: just wanted to inform you in case you didnt.
17:21:21kugeldisorganizer: I knew they conflict when I first heard that he'll edit parse_viewport
17:21:36kugelsince that function is moved to another file in my patch
17:27:05 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
17:27:21disorganizerppbly we should commit the move in a seperate patch?
17:31:24 Quit disorganizer ("ChatZilla 0.9.81 [Firefox]")
17:32:23 Join mcuelenaere [0] (
17:40:22 Join DavidS1 [0] (
17:40:41kugellinuxstb: ping, what do you think about that?
17:41:01kugelWhat disorganizer suggested
17:45:08kugellinuxstb: I also want to know if having the list_info viewport in the gui_synclist struct or as a global (which is reloaded at settings change) is preffered
17:50:29 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
17:52:20Nico_Pdoes the sansa e200 have a hard reset?
17:52:37gevaertsNico_P: hold power for a long time (15+ seconds)
17:53:13 Join Llorean [0] (
17:55:12 Join ender [0] (
17:55:45 Join mcuelenaere [0] (
17:55:53 Quit onakijmeo (Remote closed the connection)
17:57:16 Join mnakilopn [0] (i=0@
17:57:44 Join herrwaldo [0] (
17:57:59 Quit DavidSG (Read error: 110 (Connection timed out))
17:58:35 Join dimako [0] (n=546e8307@gateway/web/cgi-irc/
17:59:44dimakohow are you?
17:59:55pondlifeI'm on topic, thanks :)
18:00:00BigBambidimako: This is an on topic channel
18:00:18kugelGood, but if you have a rockbox related questio, then ask it. Otherwise join #rockbox-community
18:00:38dimakowell i was just trying to be polite :) well im having issues with the sansa e250 update....
18:00:49BigBambiUpdating rockbox?
18:01:13dimakoit seems that what ever i do the rockbox doest boot and it still the old firmware which is booting
18:01:14 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
18:01:28BigBambidimako: Have you already installed Rockbox?
18:01:40*BigBambi doesn't understand what you are upgrading
18:02:00kugelHe probably didn't install the bootloader
18:02:03dimakoyes i have tried the manual install and also the rockbox tool install
18:02:15 Nick courtc_ is now known as courtc (n=court@unaffiliated/courtc)
18:02:21BigBambikugel: I suspect so too, but "update" makes me wonder
18:02:31dimakowhen i manually installed the bootloader i have revieved a message that the install was succesfull
18:02:34kugeldimako: Did rockbox ever start?
18:02:40BigBambidimako: OK
18:02:42dimakoi ment install not update sorry
18:02:50BigBambidimako: And you extracted the build?
18:02:52dimakonope it didnt ever start :(
18:03:11dimakoyes i have extracted .rockbox to the root folder
18:03:19BigBambiDid you ever see the rockbox bootloader?
18:03:26kugeldimako: What does it say upon boot? If you installed the bootloader, some tiny text white text on black background should appear before the OF starts
18:04:07dimakono i didnt it seems to be the problem when i disconnect the sansa after the bootloader installation it just writing refreshing database and keeps working with the sansa firmware
18:04:27BigBambiAnd I assume you have turned it off then on again /stupid question
18:04:58kugelBigBambi: But he said he allready used rbutil
18:05:08BigBambikugel: So?
18:05:20kugelThat would require him clicking on Uninstall bootloader
18:05:38kugelIf he turned it off again?
18:06:07 Quit gevaerts ("work->home")
18:06:07BigBambiWhy does turning the sansa on or off have anything to do at all with intsalling or uninstalling anything?
18:06:17kugelMeh, I probably got your sentence wrong, what did you mean by "turned it off then on again"
18:06:23BigBambiEr, the sansa
18:06:36BigBambiTurn off does not mean uninstall
18:06:37dimakoi turned it off and then back on again but it seems like it boots very quick so maybe it goes into sleep mode and not really does turn off
18:06:40BigBambiIn any way
18:07:25kugelI thought you mean turn the bootloader on and off again (Yes, I read "on and off" instead of "off and on")
18:07:25BigBambiUninstall / install are completely unrelated and do not mean turn on/off in any way (or vice versa)
18:07:28dimakoi turn it off it says goodbye and then turn it on and it turns on in a second....
18:07:34 Join Tuplanol1a [0] (
18:07:41BigBambikugel: You cannot turn a bootloader on and off (or off and on)
18:07:45BigBambiNot in English
18:08:00kugelBigBambi: As I said, I got your sentence wrong
18:08:13kugeldimako: Sleep mode? I didn't know e200 have this
18:08:17dimakoby the way when i have tried now to click on install bootloader using the tool it said i already have the bootloader installed
18:08:28BigBambidimako: That is sometimes wrong
18:08:36BigBambi(fixed in SVN I think)
18:08:39dimakoi didnt say it is a sleep mode i ment that it boot very quick
18:08:46dimakotoo quick i suspect
18:08:54BigBambidimako: What version is the Sansa firmware?
18:09:04kugelAFAIK sansa of doesn't boot quickly either
18:09:07 Join mcuelenaere [0] (
18:09:35BigBambiOK, v1 :)
18:09:46BigBambidimako: So you have tried sansapatcher?
18:09:48kugeldimako: That version will not work
18:09:55BigBambikugel: eh?
18:10:01BigBambiAh, MSC only?
18:10:14kugelThis version doesn't offer MSC mode AFAIK
18:10:18BigBambigood thinking
18:10:27dimakoi have used msc and sansapatcher
18:10:31LloreanDoes the MSC trick work in it like it does on the c200s?
18:10:43dimakoyou mean i cant install the firmware? :(
18:10:43LambdaCalculus37Llorean: AFAIK, yes.
18:10:45kugelI've repeatetly heard that those 1.03.xx versions make booting into rockbox impossible
18:10:58LambdaCalculus37dimako: Have you considered a firmware downgrade?
18:11:07BigBambidimako: Even if that version is a problem, you can change the firmware
18:11:10LambdaCalculus37Some c200 users have done that.
18:11:18dimakowhy downgrade?
18:11:25 Quit ender` (Read error: 110 (Connection timed out))
18:11:29Lloreankugel: But have you heard it from people who've actually said *why* it does so, or just people who run into problems because it breaks, for example, the installer auto detect or something?
18:11:49BigBambidimako: So you are in MSC mode?
18:11:55dimakoyes i am
18:12:13kugelLlorean: I've heard that it makes installing the bootloader impossible
18:12:23Lloreankugel: That does not in any way answer what I've asked.
18:12:30kugelAnd I'm pretty sure that MSC mode isn't available on these versions
18:12:36BigBambidimako: And what is the exact message sansapatcher gives you?
18:12:46dimakoit is available and i have used it
18:13:13dimakohere is the messeage: [INFO] Bootloader installed successfully.
18:13:26BigBambiheh :)
18:13:43kugelLlorean: These people said that (regardless if rockbox was installed before) that upgrading to 1.03.xx breaks rockbox (every OF update does so) but also makes bootloader installation impossible. I don't know what you want to hear more
18:14:46linuxstbkugel: In what way does it break bootloader installation?
18:14:57kugelNo MSC mode
18:15:00dimakohey guys i succeeded it works now!
18:15:05kugelI allready told
18:15:30BigBambidimako: Cool :)
18:15:41kugelNot sure how dimako got it too work now though
18:15:52BigBambikugel: Well given dimako has 01.03.01, and he has both MSC and it worked...
18:16:04linuxstbkugel: OK, but I wouldn't call describe lack of MSC mode as making bootloader installation impossible - they just need to use the e200rpatcher method.
18:16:22dimakowell just to let you know the issue was solved buy opening the screws and removing the battery for few seconds
18:16:32kugellinuxstb: I guess they didn't have this idea (i didn't have it either)
18:16:47kugelThey intentionally tried to use e200patcher
18:16:48BigBambidimako: Hmmm, maybe it hadn't shut down properly
18:16:59BigBambidimako: Odd, but I'm glad you got it
18:18:08 Quit Tuplanolla (Read error: 110 (Connection timed out))
18:18:08 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
18:18:10dimakothank you very much for your help
18:18:26 Join ap0 [0] (
18:18:26BigBambiNo probs :)
18:18:35 Quit ap0 (Client Quit)
18:18:55dimakojust one more question if i may ask
18:20:27dimakohow do i check if my micro sdhc works now
18:20:33kugelAhh ok, I seem to have found my error. MSC isn't available in 1.03.07
18:20:47kugelI was under the impression that it's the same for all 1.03.xx
18:21:03kugeldimako: Insert it ;)
18:21:05BigBambidimako: Rockbox should read it
18:21:33BigBambidimako: But the OF still won't, so until Rockbox on Sansa gets USB you will need to use a card reader to fill it
18:21:46BigBambi(or cut and paste within Rockbox, but that is a bit fiddly)
18:22:03dimakowell how i can see the files that are on my built in memory?
18:22:09dimakoit is blank now
18:22:29kugeldimako: Maybe you should take a look at the manual
18:22:39dimakook thank you
18:22:43 Join desowin [0] (n=desowin@atheme/developer/desowin)
18:23:06BigBambidimako: The OF hides the music folder - use a different one or set show files to all in Rockbox
18:23:18BigBambiAnd use MSC to transfer files, or the Rockbox database
18:25:18 Join mcuelenaere [0] (
18:25:54 Quit krz (Read error: 104 (Connection reset by peer))
18:25:58 Join kraizee [0] (
18:27:44*Nico_P apparently can't build the manual
18:28:43Nico_PI keep getting "! LaTeX Error: Missing \begin{document}.", after which I need to type X <enter>, and I seem to get no output file
18:28:50BigBambiYou are not alone
18:28:55BigBambiI get that too
18:29:21BigBambiBut I couldn't find out why
18:29:43 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
18:30:01bluebrotherNico_P: shouldn't the bmp resize be available on greyscale targets as well?
18:30:11Nico_Pbluebrother: hi, I have a manual issue
18:30:24bluebrotherthere's album art on greyscale too, so ...
18:30:26Nico_Pbluebrother: probably but the code seems to be colour specific ATM
18:30:34bluebrotherok, what's the issue?
18:31:18bluebrother(the manual issue of course ;-)
18:31:33Nico_PI keep getting "! LaTeX Error: Missing \begin{document}.", after which I need to type X <enter>, and I seem to get no output file
18:31:40BigBambiMe too!
18:31:47dimakowell i have another problem... as soon as i connect the usb the sansa freezes and i get a message that usb didnt recognize, and i can only remove the battery to reboot the sansa
18:32:09 Join Tuplanolla [0] (
18:32:49linuxstbdimako: Hold the power button for about 10-15 seconds - that's a hardware power-off
18:33:07bluebrotherhmm. do you have a file rockbox.tex in the manual subdir of your build folder? That should contain the \begin{document}
18:33:27BigBambibluebrother: Nico_P This is what I get:
18:33:50Nico_Pbluebrother: yes, I have it
18:34:07BigBambiMe too, and \begin{document} is the third line of it
18:34:23Nico_PBigBambi: I think I get roughly the same thing
18:35:01dimakook but why cant i use the usb? how can i copy stuff to the sansa player?
18:35:12BigBambidimako: Use the OF for now
18:35:21bluebrotherinteresting. Can you try cd'ing to the manual folder and pdflatex rockbox-build.tex manually?
18:35:23BigBambiRockbox USB mode for the Sansa is being worked on
18:35:46BigBambibluebrother: Same error
18:35:48Nico_Pbluebrother: same thing
18:36:00dimakoso you mean i cant use the usb for now?
18:36:14BigBambiNot in Rockbox, boot to the Sansa firmware
18:36:21BigBambiRockbox can dual boot (see the manual)
18:36:28linuxstbdimako: You need to use the original firmware - easiest way is 1) power-off ; 2) insert usb cable ; 3) power on
18:36:41dimakooh ok thanks
18:36:46kugeldimako: Manual...
18:37:00 Quit Tuplanol1a (Read error: 110 (Connection timed out))
18:37:08 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
18:37:55bluebrotherhmm, it works fine for me. Have you tried running configure again?
18:38:10Nico_Pbluebrother: I had freshly run it to try out building
18:38:11BigBambiNico_P: What distribution?
18:38:20Nico_Pubuntu 7.10
18:38:39Nico_PI orignally wanted to try to solve the faulty unicode chars on the manual index
18:38:42BigBambiHmmm, OK - it worked for me in kubuntu 7.10, but now doesn't in kubuntu 8.04
18:39:00Nico_PI'm on kubuntu too actually
18:39:05BigBambiYeah, I was trying to write a manual patch but gave up when I couldn't build a clean manual to test it
18:39:22bluebrothercould this be an ubuntu issue (again)?
18:39:45kugelbluebrother: I build the manual successfully some month ago
18:40:04kugelNot sure how much has changed this then regarding bulding manual
18:40:11BigBambibluebrother: It built fine for me in kubuntu 7.10
18:40:14bluebrotherbuilding still works fine on my box.
18:41:03kugelbluebrother: I was on ubuntu, if it wasn't clear
18:41:39BigBambibluebrother: Can we check the versions of the required packages maybe?
18:41:41 Part codin ("Leaving")
18:41:54bluebrotherplease check for utf8x.def −− maybe it's an issue with the uft8 package?
18:41:58DerPapstgevaerts: (for the logs) my uni account is down. i'll add the comment later with a link to the patch... Or is pastebin ok?
18:42:01 Quit DerPapst ("So Long And Thanks For All The Fish!")
18:42:20bluebrotheron my box that's at /usr/share/texmf/tex/latex/unicode/utf8x.def
18:43:04BigBambihmmm, no unicode directory in ..../latex
18:43:06DavidS1In the checkwps tool what do the different verbose-ness levels mean?
18:43:12BigBambiI'll look for it elsewhere
18:43:26bluebrotherBigBambi: best ask locate ;-)
18:43:33BigBambiyep :)
18:43:57Nico_P /usr/share/texmf/tex/latex/ucs/utf8x.def or /usr/share/texmf-texlive/tex/latex/ucs/utf8x.def
18:43:59BigBambibluebrother: In /usr/share/texmf/tex/latex/ucs/utf8x.def and /usr/share/texmf-texlive/tex/latex/ucs/utf8x.def
18:44:12BigBambiNico_P: Snap!
18:44:21bluebrotherit's around two times?
18:44:33bluebrotherthat's somewhat strange.
18:44:37 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:45:03bluebrotheryou could try commenting out the \usepackage[utf8x]{inputenc}
18:45:20bluebrotherthat way you can figure if it really is an issue with the utf8 package
18:45:30bluebrotherah, forgot to mention that: in preamble.tex
18:45:44BigBambiin the manual subdir of the build dir?
18:45:47BigBambiOr in the source?
18:46:11bluebrotherin the source, as that will get copied upon make
18:46:22bluebrotheryou can just add a % in front of that line.
18:46:26Nico_Pbluebrother: I still get the same issue
18:46:40Nico_PI didn't run make clean though
18:46:59bluebrothershouldn't make a difference. Now it's really getting strange.
18:47:20 Quit kugel ("ChatZilla 0.9.81 [Firefox 3.0b5/2008032620]")
18:47:28BigBambiNo, still doesn't work after a make clean too
18:48:26*domonoky just got his new m:robe 100 :-)
18:48:37bluebrotherdomonoky: ebay?
18:48:40 Nick fxb is now known as fxb__ (
18:49:03bluebrotherBigBambi: can you have a look at features.tex? It gets created in the manual build dir and should contain several \edef lines
18:49:26BigBambiI'll pastebin it?
18:49:44bluebrotherwon't hurt.
18:50:28bluebrotherhmm, is 30GBP a good price for a MEGF20S?
18:50:54BigBambiNot too bad - I paid about £65 all in for a F40
18:50:55bluebrotherBigBambi: hmm. The first line is definitely wrong ...
18:51:02BigBambi(sorry 65 gbp)
18:51:07BigBambibluebrother: the -e?
18:51:12bluebrotherI'd prefer a bigger drive too.
18:51:13BigBambiShall I delete it and try?
18:51:39bluebrotherBigBambi: yeah, delete it and then run pdflatex rockbox-build.tex (not make, that will recreate the features file)
18:52:13BigBambibluebrother: That worked
18:52:13 Join Buschel [0] (
18:52:20BigBambiJust checking the pdf produced now
18:52:27bluebrotherI bet it's the "echo -e" line in features creation :(
18:52:46BigBambiWhich is sticking the erroneous -e in
18:53:10*bluebrother is going to check against the changes when fixing building on Ubuntu a while back ...
18:53:32BigBambibluebrother: At a cursory glance the manual produced looks OK
18:54:38*domonoky now tries the newest rbutil on his new m:robe 100 :-)
18:55:19*bluebrother wanted to trigger autodetection if an invalid configuration is found
18:55:46ByeongKeonhello domonky.
18:56:30domonokyhi ByeongKeon
18:57:40 Quit petur ("work->home")
18:57:41ByeongKeondomonoky: you told me that you'll leave some comments or questions on my application, you don't?
18:59:17ByeongKeondomonoky: did you leave something?
18:59:19bluebrotherBigBambi, Nico_P: I'll try to prepare a patch. Give me some minutes ...
18:59:26BigBambiSure, thanks :)
18:59:30domonokyByeongKeon:no, i just returned from work..
19:00:30 Join HEx [0] (
19:01:36domonokyByeongKeon: at the moment i dont have any questions for you, your application seems fine for me... :-)
19:01:58ByeongKeonThank you for review.
19:03:01ByeongKeondomonoky: I'm waiting you, for three hours. here is AM 2:00
19:03:36roolkudomonoky: let me know if the installed rockbox works - my fresh compile hung this morning, but I didn't have time to investigate
19:04:04bluebrotherroolku: I had a strange issue with the m:robe yesterday. After updating it didn't work
19:04:07domonokyroolku: i just install it, and it hangs.. :-)
19:04:15ByeongKeondomonoky: I just want to know that you still have a question.
19:04:33bluebrotheronly after several tries and ultimately removing all traces of Rockbox I got it running again.
19:04:44roolkuI suspect it is one of jhMikeS' recent changes - will try to identify which commit caused it when i get home
19:04:48domonokythe bootloader load, and then a small square of the screen goes blank, nothing else... lets reset :-)
19:05:02 Quit Mouser_X (Read error: 110 (Connection timed out))
19:05:24roolkufor me it hung after displaying the CRC in the bootloader
19:05:24bluebrotheroh, and another annoying thing: I booted with only bootloader installed. The bl didn't hang with an error message but hang in a reboot loop
19:05:37bluebrotherroolku: interesting, had the same issue.
19:06:07domonokyhm, reseting it doesnt help.. lets try a older build..
19:06:14roolkubluebrother: but you got it working with the same binary?
19:06:56bluebrotherI have r16980 running right now.
19:06:56ByeongKeondomonoky: good-bye.
19:07:04domonokyByeongKeon: i think we have more questions for you in the next few days, when the ranking of the applications starts.. so be sure to drop by here..
19:07:24roolkubluebrother: I can have a look at the bootloader behaviour, but I suspect it shares it with the other mi4 targets
19:08:30pondlifeWould it be oik if I removed the other sim debugging output calls from SVN...?
19:08:41*roolku goes home now...
19:08:50pondlifeStuff like "settings_apply()" being displayed....
19:14:38 Quit ByeongKeon ()
19:14:40***Saving seen data "./dancer.seen"
19:16:14*Nico_P is starting to hate the "reconnect mass storage" entry in the debug menu
19:16:38Nico_Ppondlife: I wouldn't mind in the least
19:16:43bluebrotherwhy? Because it's last in the list?
19:17:07Nico_Pbecause it's right next to view buffering thread and I often hit it by mistake
19:18:44bluebrotherBigBambi, Nico_P: please try this diff:
19:19:04 Join RoTtE [0] (
19:19:07 Part RoTtE
19:19:24bluebrotherit's a similar approach to the last ubuntu issue a while back. Works fine here (but no dash around)
19:19:42 Join Buschel_ [0] (
19:19:51 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:20:18 Quit Buschel (Read error: 104 (Connection reset by peer))
19:20:24pondlifeNico_P: That should be most of it.... :)
19:21:15*Nico_P updates
19:21:49pondlifeI left the file i/o stuff in by accident... oops
19:21:56pondlifeI mean, as C++ // s :)
19:21:56Nico_PI'm really loving vim's autocomplete
19:23:05BigBambibluebrother: Seems fine :)
19:23:29BigBambiI now have a nice pdf :)
19:23:34bluebrothernice, then I'm going to commit.
19:23:45BigBambiCheers for looking :)
19:24:04Nico_Ppondlife: I hope you're planning on keeping those "we open the real file..." messages commented out?
19:24:18pondlifeAs opposed to removing them?
19:24:33Nico_Pas opposed to reenabling them
19:24:55pondlifeIs // acceptible for this? I fear not....
19:25:02pondlifeI should probably /* */ them
19:25:17pondlifeOther places use //DEBUGF though
19:25:32*domonoky now has his m:robe 100 up and running with rockbox, now i have to get used to the virtual buttons and the button map :-)
19:25:52Nico_Pbluebrother: I'll try the fix when you commit it but I'm confident it will work. thanks :)
19:26:43bluebrotherdomonoky: the buttons are a bit ... inconsistent on the m:robe. Wanted to create a patch for that, but ...
19:27:02domonokywhat does this "Display" button do ?
19:27:45 Quit dimako ("CGI:IRC")
19:27:49bluebrotherNico_P: committed.
19:28:03*Nico_P has an improvement to audio_current_track that doesn't need to spin the disk
19:28:26pondlifeNico_P: Sounds good for battery life...
19:28:28 Nick Bagder_ is now known as Bagder (
19:28:29bluebrotherthanks guys for noting −− I haven't introduced the features feature but it's always better to have it building ;-)
19:28:46bluebrotherdomonoky: some awkward things −− resume from the browser for example
19:28:50*amiconn would like to see the greylib ported to the m:robe 100
19:29:33Nico_Ppondlife: well the current one only spins the disk when the current track is unbuffered, so no big difference there. the bonus is that the disk isn't spun in the UI thread, thus not blocking it anymore
19:29:41amiconnI don't think it's much work - but doing it without a target would be hard
19:30:05 Join Buschel [0] (
19:31:12Nico_Pbluebrother: awesome :)
19:31:14domonokyamiconn: so mainly calibrating is needed ?
19:31:56*Nico_P should add an AA section to the manual
19:32:06 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
19:32:43amiconnThe first thing would be to measure, and if necesary, optimise the lcd driver. Next step would be implementing lcd_blit_mono() and lcd_blit_grey_phase(). Then measure lcd scan rate, put this into the greylib, and then finally calibrate
19:33:47markunSearching in the forums gives me a "Database Error". Does anyone know what's going on?
19:33:56amiconnAn interesting value would be the value of LCD1_CONTROL
19:34:16amiconnThat should give some hints as to whether optimisation is needed and/or possible
19:34:20 Join petur [50] (n=petur@rockbox/developer/petur)
19:34:35amiconnAnother useful thing would be running test_fps, both unboosted and boosted
19:35:14 Join ap0 [0] (
19:35:25amiconnI think I could do it remotely, but I'm not sure how many round-trips it will take...
19:35:43 Join MethoS- [0] (
19:36:11*domonoky is happy to help with this, so i will now try to compile the test_fps plugin :-)
19:37:38amiconnYou could put LCD1_CONTROL output into the ports debug screen as well
19:38:32 Quit Buschel_ (Read error: 110 (Connection timed out))
19:40:05amiconnThe good thing on the m:robe is that it's 1-bit, so there's half as much data to move per frame as on the 1st..4th Gen iPods, while using the same bridge
19:40:06domonokywill do..
19:42:07*domonoky finds it strange that the buttons on m:robe react before you touch it.. :-)
19:42:19amiconnMagic buttons ;)
19:43:57amiconnNico_P: Playlist indices on hwcodec now *are* wrong when shuffle is enabled :(
19:44:11Nico_Pdamn :(
19:44:27Nico_Pamiconn: do you have an idea why?
19:45:51amiconnHappens only when global shuffle is neabled, not when using "insert shuffled" on a dir
19:46:09Nico_Pmarkun: the forums are flaky these days
19:46:35Nico_Pamiconn: I'll see if I get it on the sim
19:46:57amiconnBut when using the latter, another quite strange effect happens when skipping: The displayed playlist index changes to 1 for a fraction of a second
19:47:32amiconnI.e. if you're on track 5 and skip forward, the display goes 5->1->6 within half a second or so
19:47:56Nico_Pthat's strange
19:48:07Nico_Pit means id3.index is 0
19:48:07roolkuamiconn: I only need lcd_grey_data() but I am having trouble understanding what the assembly versions...need to do more reading
19:48:39amiconnroolku: Did you measure fps already and/or read the value of LCD1_CONTROL?
19:49:02roolkuamiconn: measured 51.1 (I think)
19:49:02amiconnIirc the driver doesn't set it, so it should contain whatever the olympus loader sets it to
19:49:10amiconnThat's too slow for the greylib
19:49:50 Join webguestHdaackda [0] (n=dd78d126@gateway/web/cgi-irc/
19:50:01webguestHdaackdahello. anyone there?
19:50:06amiconnYou'll need at least twice the internal scanrate of the lcd, which is usually around 75Hz (e.g. archos: 67Hz, H1x0: 70Hz, iPod 2nd Gen: 95Hz etc)
19:50:36amiconnThis can be achieved by programming LCD1_CONTROL, but first I need to know the current value, in order to know the mode used in the m:robe
19:50:37ali_asThere is noone here, at all. Just chickens.
19:50:51webguestHdaackdaah good. i need to ask something :). there used to be a GUI selector for ipods (meaning u could select apple, rockbox etc..)
19:51:32roolkuI'll let you know as soon as I get it to boot again - a recent commit broke it
19:51:35webguestHdaackdaas in, we had a choice what to load when ipod was restarted.. any idea whats it called, how to install it etc
19:53:33ali_asSomeone will know, patience.
19:53:55bluebrotherwebguestHdaackda: no, there was never.
19:54:00webguestHdaackdak.. if its any help, i think it had to do something with ipod liinux port too
19:54:01amiconnNico_P: Here's the really old fix for the same problem, from archos-only times:
19:54:13bluebrotheryou're referring to the bootloader. There is Loader2 of IPL around which offers a menu.
19:54:28bluebrotherRockbox bootloader does not include a menu (and probably will never).
19:54:32domonokyamiconn: the newest rockbox build doesnt boot on my m:robe so i cant test :-/
19:54:37amiconnNote the simple use of id3->index + 1 before, and the calculation in the new revision
19:54:43bluebrotheryou can use IPL's Loader2 to load Rockbox, but that is not supported here.
19:54:49J-23Can I run Rockbox VMware Development Platform on real computer from pendrive/memory card/any other mass storage device?
19:55:19bluebrotherJ-23: I wouldn't suggest using a flash memory drive, but apart from that, yes.
19:55:30bluebrotherit's just a file vmware uses ...
19:55:38amiconnYou need to set a sticky revision in viewvc in order to get to the old wps.[ch]/wps-display.[ch] files
19:55:48webguestHdaackdayeah, thats it. loader2.. but tht did hav some problems.. rockbox.ipod has to be on the root and not inside rokbox for that to work
19:56:20 Join low_light [0] (i=c730190b@gateway/web/ajax/
19:56:37Nico_Pamiconn: thanks. any idea why the index becomes 0 at one point?
19:56:44bluebrotherwebguestHdaackda: well, Loader2 is not supported by us.
19:57:10amiconnI don't really understand how the engine calculates the indices in global shuffle mode
19:57:15LambdaCalculus37Loader2 questions are something you have to ask the iPodLinux guys about.
19:57:40amiconnAs long as global shuffle is disabled, the inidces correctly reflect playlist position
19:58:01 Join phinze [0] (
19:59:09 Join lee-qid [0] (
19:59:49 Quit ap0 ("Bye.")
20:00:10webguestHdaackdathanks bluebrother
20:00:40J-23bluebrother: Do I need convert rockbox-flat.vmdk to raw disk image with qemu-img?
20:01:37low_lightroolku: hi...suspect that the problem with the m:robe is the bss size is not properly aligned due to the recent changes to
20:02:05bluebrotherJ-23: it's a vmware image, just run it in vmware. No idea about other emulators.
20:02:06low_lightI had that problem early on, but something along the way cleared it up.
20:02:06 Quit webguestHdaackda ("CGI:IRC (EOF)")
20:02:41 Quit nikosapi_ (Remote closed the connection)
20:02:44 Join gevaerts [0] (
20:03:00low_lightroolku: you could try adding ". = ALIGN(0x4);" the line before _iramend in firmware/target/arm/olympus/
20:03:19roolkulow_light: hm, I'll have a look - but I can't say I know what I am doing. :)
20:03:28roolkuwill do
20:04:12amiconnYOu could also simply check the .map whether alignment is correct
20:04:49low_lightI see that when I commited the port that bit of code ended up in the tms320dm320 section by mistake (oops)
20:05:35amiconnm:robe .lds should be identical to any other pp502x .lds file
20:05:48amiconnIf it's not, something is clearly wrong
20:06:12*amiconn builds an mrobe build even though he has no mrobe
20:08:09roolkuit is identical to the h10 one
20:08:10*low_light waits
20:08:41*roolku started a build before getting a chance of looking at map files
20:08:41 Quit bluebrother (Read error: 104 (Connection reset by peer))
20:08:54 Join bluebrother [0] (
20:10:03amiconnWell, you need to have a build in order to have a .map of it...
20:10:04roolkulow_light: what bit of code are you referring to (tms320dm320 section)?
20:10:14 Join stripwax [0] (
20:10:35roolkuamiconn: yes, but make clean wiped the one from the previous compile
20:11:00*amiconn does usually build without a prior 'make clean'
20:11:21amiconnDoing that would really be annoying on cygwin...
20:11:38low_lightroolku: the align statement right before iramend
20:11:41roolkume too, but I have done various updates to specific revision, that usually confuses make
20:13:08amiconnUmm, .bss is indeed unaligned...
20:13:21roolkuthe align solved indeed solved the problem
20:13:31amiconnHmm, but it's auto-aligned within the section
20:13:48amiconnMaybe not all linker versions do this?
20:14:03amiconn$ arm-elf-ld -v
20:14:03amiconnGNU ld version 2.16.1
20:14:36 Join ender` [0] (
20:14:40 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
20:14:44amiconnEh, I know what happens...
20:15:16amiconnThe section _content_ is auto-algined, but since the _section_ is unaligned, the bss zero-out in crt0.S will crash...
20:15:22roolku$ arm-elf-ld -v
20:15:23roolkuGNU ld version 2.16
20:15:34 Quit ender (Read error: 104 (Connection reset by peer))
20:15:51amiconnImo it's a good idea to add this alignment fix to all arm targets
20:16:18roolkuwant me to do it?
20:16:22 Join Massa [0] (
20:16:24J-23Can I copy Rockbox VMware Development Platform HDD image to any directory on my disk, add it to another image and modify to run it on real computers?
20:16:37amiconnWe were just lucky that it didn't hit us on the other targets so far (luck == iram section size being a multiple of 4)
20:16:41roolkuI am suprised it only affected the m:robe
20:17:11amiconnProbably other arm targets don't put byte/short data in iram
20:18:13low_lightwhy did the last changes trigger it?
20:18:23 Nick DavidS1 is now known as DavidSG (
20:20:48amiconnIt's a sub-section order problem in the iram section
20:21:20amiconnm:robe has the .idata part of rolo.o last, which is only 2 bytes
20:21:28pondlifeMe doesn't quite understand why playlist_resume() is required... resume is a playback thing, not a playlist thing, surely.
20:22:32amiconnBut e.g. mini G2 has .idata parts of other objects after that, so the linker auto-aligns it
20:26:00amiconnpondlife: It's both, as the resume postion consists of the playlist index and the resume point within the track
20:27:35amiconnNico_P: On swcodec your fix seems to work nicely. So we could ifdef for swcodec/hwcodec, but somehow I think that wouldn't be right...
20:27:49amiconnCould be a quick intermediate fix though
20:29:26 Quit phinze ()
20:32:35*domonoky has just run test_fps on his m:robe : 1/1 -> 119.6 fps 1/4-> 469fps greyscale 1/1 -> 281fps .. all at 30 Mhz..
20:34:07roolkudomonoky: try the scanrate plugin :)
20:34:14 Quit Rob2222 ()
20:34:24domonokyat 80Mhz: 1/1 319fps 1/4 -> 1251.5fps, greyscale 1/1 280 fps..
20:34:34LambdaCalculus37domonoky: What are your measurements when boosted on the m:robe?
20:34:42*LambdaCalculus37 should be a wee bit faster in asking :)
20:36:29 Join Rob2222 [0] (
20:36:40roolkuamiconn: LCD1_CONTROL is 1685 or 0x695
20:37:27MassaNico_P: Hi - it seems you're here every day - don't you have a real life ;)
20:37:47Nico_PMassa: hehe, it's not too busy ATM
20:38:03Nico_PI do have an exam tomorrow though
20:38:07roolkuhm, shouldn't this be a byte...
20:38:24Nico_Pamiconn: I think I'll go with the ifdef for now
20:38:36MassaNico_P: so you should learn or do you already know everything? :O
20:39:01 Quit linuxstb (Nick collision from services.)
20:39:04Nico_PMassa: I still need to work, yes
20:39:14 Join linuxstb [0] (
20:39:40MassaNico_P: so I'll ask my questions now and not later ;): Did you think about the optional parameters in parse_list? I have some suggestions
20:39:51 Join disorganizer [0] (
20:40:00Nico_Pno, I didn't think about it, sorry
20:40:01low_lightroolku, amiconn: LCD1_CONTROL is explicitly set in lcd_init (same as the OF bootloader)
20:41:07MassaNico_P: I thought about two things: first a leading additional character (e.g. a '+') which shows that the parameter is optional (but needs separators around it)
20:41:34roolkuso it is and to several values as well. :) 0x694 being the last
20:41:46MassaNico_P: and second a special character (e.g. a '|') which shows that all following parameters are optional (and don't need separators).
20:42:16low_lightamiconn, roolku: perhaps the optimizations in r15655 also apply to the m:robe?
20:42:41Nico_PMassa: I like the second idea better, it sounds less complicated
20:42:58MassaNico_P: so with e.g. "+c" you could say the the colour definition could be there or could also be "empty" (two separator chars)
20:43:20MassaNico_P: I thought to use both ideas...
20:43:24amiconnlow_light: I would think so, yes
20:43:38Nico_PMassa: maybe, yes
20:44:10MassaNico_P: so you're not sure if it's necessary?
20:45:18amiconnBit 0 should have no effect. Bit 2 is the controller reset (0 == reset), so lines 149/150 reset the lcd controller
20:45:42MassaNico_P: Another (more complicated) version would be a character which also says "the parameter may be empty" and additionally says "and the next argument in the va_list is the default value which is used if the parameter in the string is empty...
20:46:10 Join Mouser_X [0] (
20:46:10Massawould be more flexible but also harder to understand...
20:46:11amiconnBit 4 and 7 probably need to be set on mr100 (on the ipods, bit 4 must not be set)
20:46:53amiconnBits 9 and 10 determine the speed. So try setting it to 0x94 instead of ox694 - that would be the fastest possible setting
20:47:07Nico_PMassa: yeah I think that's going too far... but I actually do like the "+ and |" idea
20:47:31pixelmaMassa: I don't like more chars one needs to "escape" much (just an aspect of it though)
20:47:33amiconnIf that's too fast for the lcd controller, try 0x294, and if that's still too fast, 0x494
20:48:30amiconn319fps @80MHz would be sufficient even if the controller doesn't like the faster setting, but faster == better here
20:48:42domonoky5 hours left till gsoc deadline.. students hurry up ! :-)
20:48:58Massapixelma, Nico_P: we could also just use the uppercase version of the characters to say "the parameter may be empty"...
20:48:59Nico_Ppixelma: it's not meant to be a user visible change
20:49:18gevaerts5 hours and 10 minutes. Thise are students, so the last 10 minutes matter :-)
20:50:01MassaNico_P: without default values what will be given back if a parameter is empty?
20:50:10amiconnlow_light: I think lcd_init_device() for mr100 can be shortened quite a bit
20:50:24roolkuunboosted: 381.5 1487.6 217.0 fps
20:50:44amiconnWith 0x94 ?
20:50:52MassaNico_P: int: -1, strings: null, hex colour -1?, greyscale -1?
20:51:09amiconnThe greyscale speed is meaningless atm
20:51:48amiconnTest whether it still works when boosted
20:52:00amiconnThe greylib needs to boost anyway
20:52:08amiconn(on PP)
20:52:21roolkuboosted: 1018.0 3971.5 214.5 fps
20:52:55low_lightamiconn: quite possibly. it's basically copy/paste from the of bootloader.
20:52:56Nico_PMassa: I would think more like int:0, strings:"", coolor:black
20:53:11*roolku has a distorted greyscale cube :)
20:53:18amiconnWow, the first target exceeding 1000fps :)
20:54:07amiconnNo real surprise though, as the fullsize ipods reach 540fps (twice as much data to transfer)
20:54:10MassaNico_P: and how would the caller distinguish about a "given" 0 and the "default" 0?
20:54:14roolkuI think my lcd_grey_data is slightly wrong (stole it mainly from ipos)
20:54:45amiconnroolku: The ipod version won't work because of the different pixel packing, but no worries, I'll provide one
20:54:55Nico_PMassa: would it need to? if so, your values are probably better
20:55:13ChristopherWI found a bug in the file browser's context menu, but I don't know which category I should put it in in Flyspray; should it be "Applications", "User Interface", or something else?
20:55:39linuxstbChristopherW: What's the bug?
20:55:39roolkuit looks more like a rhomboid (if there is such a thing)
20:55:51amiconnImo it should also do inversion on mr100, in order to not see photographical negatives
20:56:16 Join Spex [0] (
20:56:20roolkuamiconn: I think it would be better to just to a hardware inversion for photo-type applications
20:56:31ChristopherWlinuxstb: when a directory is empty, the context menu still shows most file operation items (Cut, Copy, Properties, etc)
20:56:48amiconnroolku: Inversion is free here.
20:56:53roolkuamiconn: the cube looks fine as is (and I imagine strange as dark on red background)
20:57:20ChristopherWit also has a Rename context menu item to rename a file that isn't there
20:57:28amiconnThe greylib can be shown together with normal display content, so hardware inversion wouldn't be nice
20:57:49linuxstbChristopherW: I guess User Interface will be fine. I'm not sure what "Applications" would be...
20:57:54 Join spiorf [0] (n=spiorf@
20:58:00amiconnroolku: Then the cube colours should be defined differently for m:robe, not the greylib made inverting stuff
20:58:08roolkuokay, point taken
20:58:23 Part pondlife
20:58:36ChristopherWI guess almost all of the stuff running in Rockbox is technically an "Application"
21:00:10roolkuscanrate is still at 51.1Hz
21:00:38Nico_Pamiconn: actually I think the ifdef is probably the best thing to do... HWCODEC can safely rely on the playlist index
21:00:56Nico_PI'll add a comment though
21:01:10amiconnroolku: Oh, you meant scanrate...
21:02:10amiconnI thought those 51.1 Hz would be the update fps...
21:04:57amiconnWell, that means the greylib will cause a quite low CPU load when running. It also menas that the display might look a little grainy, depending on the slowness of the lcd panel
21:05:12amiconn(slow panel == good for the greylib)
21:07:31 Join mirak [0] (
21:08:15 Join mcuelenaere [0] (
21:08:55 Quit mirak (SendQ exceeded)
21:09:23MassaNico_P: well, if one default value for a specific type is enough - we could simplify it without default values; but if different situations need different default values the caller either has to provide the situation dependent default or he needs to know if a default shall be used....
21:09:44*amiconn is working on lcd-as-mr100.S
21:09:55 Join mirak [0] (
21:09:58*roolku gets excited and downloads elephantsdream
21:10:01 Join stripwax_ [0] (
21:10:22Nico_PMassa: default values don't need to be given to parse_list. the caller can simply set them before calling parse_list and parse_list won't change them
21:10:44*gevaerts wonders if the red mrobe screen will make that pink elephants dream
21:11:00MassaNico_P: you're right! That would be the best solution!
21:11:08roolkugevaerts: :)
21:11:57Nico_Pamiconn: I committed the ifdef workaround
21:12:50gevaertsroolku: make sure to let us know :)
21:12:51 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
21:14:43***Saving seen data "./dancer.seen"
21:14:59preglowall this pink elephants talk makes me want a beer
21:15:57 Join tvelocity_ [0] (
21:16:01*Massa thinks that preglow already is drinking his beer ;)
21:16:22preglowi wish it was so
21:16:29*ender` yawns
21:17:12*domonoky thinks the m:robe button/touch driver has a small bug.. if i hover my finger over up or down and let it go into button repeat, i can sometimes take my finger away without stopping the repeat..
21:18:33roolkudomonoky: yes, the bar is slightly fishy still - it is not meant to be used as buttons
21:20:12 Quit stripwax (Read error: 110 (Connection timed out))
21:20:16 Join mcuelenaere [0] (
21:20:46 Join Lear [0] (i=chatzill@rockbox/developer/lear)
21:21:07 Join Ravendug [0] (
21:22:07 Join rob-e [0] (n=rob@
21:22:54MassaHi rob-e!
21:23:06rob-equick question, how do you go about loading music onto the player, is it drag and drop onto the root of the drive or what?
21:23:20Horschtdrag and drop anywhere you want
21:23:23bluebrotherit's m:robe, not rob-e *g*
21:23:39domonokyrob-e: just drag&drop where you want..
21:24:16rob-ecool, so it would be a good idea to make a music folder on the root or whatever and it reads it all
21:24:25BigBambiAs you wish
21:24:50rob-eso i could plug a ipod into a computer and play directly from it too?
21:25:18 Quit stripwax_ ("Miranda IM! Smaller, Faster, Easier.")
21:25:18BigBambirob-e: If you don't use itunes or an other music management software, yes
21:25:19*Buschel is fighting with the asm-routine in pcm-pp.c
21:25:24Buschela real bitch
21:25:30 Quit tvelocity (Read error: 110 (Connection timed out))
21:25:32rob-ethis is good, because my friend has a car stero with a usb and o man sweet
21:25:58rob-escrew itunes
21:26:38Horschtyes, screw it. big time
21:26:47 Join DavidS1 [0] (
21:26:48 Quit toffe82 (Read error: 104 (Connection reset by peer))
21:27:19 Join MethoS-- [0] (
21:27:48 Join toffe82 [0] (
21:28:50rob-ecan i pot this on a non ipod video and play videos?
21:29:27LambdaCalculus37rob-e: You could.
21:29:51LambdaCalculus37If you want to. :)
21:30:10amiconnroolku, low_light, domonoky:">
21:30:19*LambdaCalculus37 has a 4G iPod color that plays video, thanks to Rockbox
21:31:25amiconnRockbox can play video on all its supported targets, except the Archos Player (charcell LCD), and the m:robe100 (just about to fix that)
21:32:05*gevaerts adds one item to the list of possibilities that would be created by usb host support. You could use your H300 or gigabeat as a mobile rockbox installation device
21:32:16LambdaCalculus37amiconn: How will video work on the m:robe 100? Just curious.
21:32:32amiconnSame way as on all greyscale targets... using the greylib
21:32:55gevaertsBut called pinklib in this case
21:32:57*gevaerts ducks
21:33:06amiconnMore like redlib...
21:33:27bluebrotherhmm ... should we call it Redbox in that case? ;-)
21:34:02*BigBambi will go mental if more plugins appear that are either Rock* or *box :)
21:34:11LambdaCalculus37For some reason, I was thinking of RVF support, like on the Archos devices (minus the Player, of course).
21:34:11domonokyits redscale on m:robe.. :-)
21:34:55bluebrother#ifdef HAVE_REDSCALE
21:35:02 Part mcflow
21:35:26bluebrotherhmm, the redlib patch doesn't build for me :/
21:35:33*roolku twiddles his thumbs
21:35:45 Join ap0 [0] (
21:35:54bluebrotheram I missing something?
21:35:54amiconnLambdaCalculus37: Actually rvf works the same way, just the frames are precaclulated instead of being calculated on-the-fly
21:36:07 Quit ap0 (Client Quit)
21:36:15gevaertsBigBambi: if we ever get an rbutil port to rockbox going (once usb host is there), can we then get an exception and call it something with rock ?
21:36:16amiconnbluebrother: lemme check
21:36:28BigBambigevaerts: I'll think about it
21:36:30domonokybluebrother: theres a b too much in lcd-mr100.c
21:36:39amiconnI actually did just copy, paste & shuffle-around. Should have tried building it...
21:36:52domonokybwidth instead of width at line 214
21:37:23amiconnah, of course
21:37:34amiconnThere's also a declaraion missing
21:37:56LambdaCalculus37amiconn: Thanks for that bit of information. I was curious about how rvf worked. :)
21:39:27amiconnLambdaCalculus37: Nowadays we could also use the greylib because it's fast enough, but using precalculated 1-bit frames @67Hz actually saves space versus e.g. 8-bit frames @25Hz (which would then be converted on-the-fly to 1-bit frames at 67Hz)
21:39:39roolkuvoid lcd_grey_data(unsigned char *values, unsigned char *phases, int count);
21:39:53 Quit kraizee (Read error: 110 (Connection timed out))
21:39:57gevaertsmarkun: I noticed that the usb connector on the M6SL appears to be an OTG connector, not just a device-side one
21:41:11 Join Mathiasdm [0] (
21:42:22BHSPitMonkeyOkay, my app is in.
21:42:36BigBambiGSoC for Rockbox?
21:42:38LambdaCalculus37amiconn: That does make sense, actually.
21:42:43*gevaerts looks
21:42:45BHSPitMonkeyIf anyone would like to provide any input on it before the deadline, I would be appreciative.
21:42:49 Join mcflo1 [0] (
21:42:53 Part Spex
21:43:03 Part mcflo1
21:43:09Nico_PBHSPitMonkey: I believe you can still edit the app after the deadline
21:43:21amiconnroolku, low_light, domonoky: Patch updated, although I doubt you'll need it...
21:43:22 Part J-23
21:43:57 Quit MethoS- (Read error: 110 (Connection timed out))
21:44:01BHSPitMonkeyNico_P, my interpretation of the guidelines led me to believe the contrary, but I'll double check.
21:44:30gevaertsBHSPitMonkey: as I understand it you will only be able to edit in response to a comment
21:46:04bluebrotherseems lines of about 8px or so are upside down
21:46:45amiconnAh, then I got the packing order wrong...
21:47:21 Quit DavidSG (Read error: 110 (Connection timed out))
21:47:35 Nick DavidS1 is now known as DavidSG (
21:48:02markungevaerts: great!
21:48:29gevaertsmarkun: of course, this comes without any guarantee that it actually does OTG...
21:49:06amiconnroolku, low_light, domonoky: Patch updated again
21:49:18amiconnEh, and bluebrother of course :)
21:49:24markungevaerts: any quick way to find out?
21:49:25roolkuis it only the asm that changed?
21:49:30BHSPitMonkey(I have to be off to a lab now, but I'll still be receiving comments made toward me in here.)
21:49:42markungevaerts: would there have to be 5V on one of the pins for example?
21:50:00gevaertsmarkun: I haven't studied OTG specs yet, so I don't know
21:50:02MassaIs checkwps currently broken (and not compilable) or am I too dumb to do it?
21:50:17amiconnThe constants in the 8x biceq changed. They now go 0x80..0x01 instead of 0x01..0x80
21:50:37MassaI do "make MODEL=h300 TARGET=IRIVER_H300 checkwps.h300" to compile a iRiver h300 version
21:51:04*bluebrother remembers he updated the checkwps Makefile to work on MinGW the weekend
21:51:08Massaand get ../../firmware/export/debug.h:26: error: parse error before "ATTRIBUTE_PRINTF"
21:51:19amiconnIf that works, the next step would be the calibration, for which you need to compile test_grey.c (might need button definitions for m:robe)
21:52:04Massabluebrother: locally on your computer or did you check it in? BTW, I currently use cygwin...
21:52:12 Join Spex [0] (
21:52:18bluebrotherMassa: locally. Haven't checked in yet ...
21:52:34Massabluebrother: so what did you change?
21:52:37bluebrotherthe only remaining issue is the script. Doesn't run on MinGW (no sh.exe around)
21:53:12bluebrotherbasically, you need to set CC correctly. But I also reused the MODEL variable, so you now simply use the target checkwps
21:53:35Massabluebrother: that's not my problem - I directly call make (BTW, what about MSYS together with MingW?)
21:54:07bluebrotherI haven't looked into msys at all −− I just used the MinGW that was installed together with Qt
21:54:22*bluebrother watches a jpg image on m:robe100 :)
21:54:24MassaI wonder why the checkwps (for the current target) is not part of "make tools" in the build directory?
21:55:11amiconnbluebrother: Calibration will (hopefully) improve quality (usually the lcd isn't linear)
21:55:12Nico_Pbluebrother: are you aware of the unicode problems on the HTML manual front page?
21:56:20 Join sarixe [0] (
21:56:37 Part Spex
21:56:41bluebrotherNico_P: no. What characters are messed up? I sometimes had wrong characters due to some Firefox issues
21:57:02Nico_Pbluebrother: hmm I'm not getting them anymore... for some reason firefox saw the page was ISO-8859-1
21:57:41bluebrotherI was surprised when I saw that first but figured it wasn't a problem with the html itself
21:58:47bluebrotheramiconn: nice :)
21:59:01 Quit LambdaCalculus37 ("CGI:IRC 0.5.9 (2006/06/06)")
21:59:23domonokybluebrother: so amicons patch works for you ? i only get a blank screen.. (may be i messed up the patching)
21:59:52bluebrotheryep, works fine here.
22:00:10bluebrotherI just reverted the previous patch and applied the latest.
22:00:16*domonoky detects he messed up.. :-)
22:01:17 Quit Thundercloud (Read error: 104 (Connection reset by peer))
22:01:48 Join Thundercloud [0] (
22:01:57domonokyis there any documentation whats in the absolute data packet of m:robe100s touchpad ?
22:02:16amiconnSO, who will calibrate?
22:02:43amiconnIn fact it would be good to have data from several m:robes, and use the average
22:02:56bluebrothercan I do anything to help calibrating?
22:03:15amiconnYes. First you need to build test-grey.rock
22:03:19gevaertsmarkun: it's clearly a mini-AB connector. According to the usb otg spec, that means it _must_ be a dual-role device (i.e. both device and host). Of course they might still violate the spec
22:03:27amiconnIt might need button definitions for m:robe
22:03:50amiconn(Left/ Right/ Up/ Down/ OK and Quit)
22:04:11bluebrotherI already compiled that. Seems the buttons are already defined.
22:04:24amiconnThen run it. It will show 9 rectangles
22:04:30roolkuif bluebrother is doing the calibration I don't need to bother. :)
22:04:52amiconnThe top, bottom, left and right rectangle will show you the "destination" level, rastered
22:05:08amiconnroolku:[22:02:43] <amiconn> In fact it would be good to have data from several m:robes, and use the average
22:06:28amiconnThe top left and bottom right will show one step darker, the bottom left and top right one step lighter, for comparison
22:06:37bluebrotherhmm, the left button sometimes seems to act as select on my mr.
22:06:51amiconnThe center rectangle is what you need to adjust with up & down
22:07:12amiconnThere is a total of 15 greylevels, plus pure black & white
22:07:17amiconn(red) ;)
22:07:37markungevaerts: some talk about USB OTG and an external battery pack:
22:07:40toffe82gevaerts: the mini AB is 5 pins ?
22:07:43amiconnIf you have adjusted all of them to match as closely the destination brightness, press select
22:08:15gevaertstoffe82: unless I miscount, there seem to be 5 pins
22:08:22low_lightdomonoky: the documentation is on Symantec's website.
22:08:27amiconnA text file will be created in the root, containing the raw data, which I then put into Openoffice Calc for averaging and interpolation
22:08:28bluebrotherhmm, up and down buttons seem to have no effect. Checking source ...
22:08:58amiconnThe steps are rather fine...
22:09:01roolkubluebrother: that is probably your thumb hovering over the center of the strip
22:09:21low_lightdomonoky: make that
22:09:39toffe82gevaerts: from the schematics of the meizu one of the pin is not connected
22:09:40domonokyroolku: this pdf ?
22:10:02 Quit disorganizer ("ChatZilla 0.9.81 [Firefox]")
22:10:07 Quit Massa ("ChatZilla 0.9.81 [Firefox]")
22:10:26low_lightthe one's here:
22:10:46amiconnIdeally, you look at the display exactly perpendicular when calibrating, and from as far away as possible (you still need to be able to operate the m:robe ;) )
22:10:51toffe82low_light: any luck with the sa9200 ?
22:10:54low_lightI think it's the 1-d sensor document
22:11:19*amiconn should probably put a description of test-grey.rock into the wiki
22:11:44roolkudomonoky: I think this one if you want the strip:
22:11:55domonokythanxs found it..
22:12:32low_lightdomonoky: there are some options to enable that would have simplified some things, but unfortunately they don't seem applicable to the revision in the m:robe
22:12:44gevaertsmarkun: great thread :) Lots of people saying what they think might be the case...
22:13:00bluebrotherdo I need to run the plugin for each new greyscale or will all be put into a single file?
22:13:09 Join everyday [0] (
22:13:10markungevaerts: not the first thread of that kind
22:13:29amiconnbluebrother: It puts all values into a single file
22:13:33low_lightThe simpliest things to do are use the lower sampling rate. Also filter out non-presses (hovering).
22:13:49amiconn...and it also doesn't overwrite data from previous runs (uses create_numbered_filename())
22:14:32 Quit everyday (Client Quit)
22:14:37low_lightI've been playing around with incorporating making "tap" and "slide" work
22:14:43gevaertstoffe82: sounds like the only thing missing could be VBus. Maybe some kind of adapter is possible
22:14:43amiconnAnother small tip: If just go through all 16 levels at the start, you'll probably notice that most are either too dark, or most are too bright
22:14:45 Join everyday [0] (
22:15:18toffe82gevaerts: on the schematics, on the plug they call it ID
22:15:41amiconnThe tip is to start at the end to where the levels are shiften, because the plugin automagically corrects discontinuities (a darker level can never be brighter than a brighter level), so you'll have to do less button presses
22:16:21domonokylow_light: to make the strip less sensitive to hovering of the finger we could use the Z values in the packet i think..
22:16:38bluebrotherah. Will retry with that.
22:16:43gevaertstoffe82: are you good at electrical things ?
22:16:49low_lighttoffe82: nothing new. Still need the bootloader. I was hoping there would be a "manufaturers mode" like the Sansas, in which case we could use a modified version of e200tool to extract the bootloader.
22:16:52amiconnYou can of course go back & forth through all levels as often as you want, before the final Select button press
22:17:06everydayIs there any way to reduce the power draw of the MPEG plugin
22:17:27Bagdereveryday: sure, optimize it!
22:17:28bluebrotherbesides, it would be nice to have some kind of feedback when levels are changed or the current value is saved. Due to the non-mechanical feedback of the touchpad this is a bit guessing
22:17:53amiconnYOu could use button click, I presume
22:17:55davenoptimize it, make it use the dsp somehow
22:18:06BigBambiWhich dsp?
22:18:14Bagderdaven: very few targets have dsp
22:18:16amiconnwhich target?
22:18:21low_lightdomonoky: yes. I did not add that at first because there is an option to ignore Z=0 presses, but like I said it doesn't seem applicable for this revision hardware
22:18:30toffe82gevaerts: I will not say good by I understand :)
22:19:08everydayHow do I optimize it?
22:19:08gevaertstoffe82: could you have a look at, section 4.3 ?
22:19:20*domonoky thinks that Z is higher than zero, when we have this hovering effect, will experiment tomorrow a bit... :-)
22:19:26Bagdereveryday: if you need to ask, you're not the man to do it...
22:19:27amiconntoffe82: The ID pin isn't Vbus. The ID pin is used to distinguish a mini-A from a mini-B plug in a mini-AB socket
22:19:41BigBambieveryday: Dive into the code, and make things happen in less cycles
22:19:44amiconn(one of them shorts ID to ground, iirc it's mini-A that does that)
22:20:25gevaertsamiconn: you do rc :)
22:20:25 Quit everyday ("ChatZilla 0.9.81 [Firefox]")
22:20:50 Join parafin|away [0] (
22:20:55 Join dabujo_ [0] (
22:21:00Bagderhe couldn't wait to start optimizing I guess
22:21:01amiconnThen there is the very special iaudio M5, which mis-uses the ID pin...
22:21:26low_lightdomonoky: there is also a sensitivity adjustment which could probably be adjusted to eliminate most of the hover effect
22:21:27 Quit parafin (Nick collision from services.)
22:21:31 Nick parafin|away is now known as parafin (
22:23:12toffe82amiconn , gevaerts : on the schematic of the meizu the ID pin is not connected (markun gave me the link to this schematic)
22:23:50markunbut those are the schematics for the M6SP, it might be different in the M6SL
22:24:03gevaertstoffe82: from my reading of that paragraph, that would mean it shouldn't have an AB receptacle ...
22:24:07amiconnThen it cannot support OTG mode, but if it's an OTG controller, it could be manually switched to host mode
22:24:29toffe82gevaerts: looks like a mini B
22:24:37amiconnBut you'd need a special adapter cable, as a mini-A plug won't fit into a pure mini-B socket
22:24:38 Join stripwax [0] (
22:24:59 Join MethoS- [0] (
22:25:11gevaertstoffe82: the 'not connected' there is for the cable I think
22:25:34gevaertsamiconn: cable won't be a problem. It's a mini-AB socket
22:25:47amiconnbluebrother: Ah :) A quite typical lcd curve from a first look
22:25:58 Join csc` [0] (n=csc@archlinux/user/csc)
22:26:11toffe82gevaerts: if you look from the player, ID is not connected so when it is connected to a computer for example it must be seen as a mini b device
22:27:14 Join shotofadds [0] (n=rob@rockbox/developer/shotofadds)
22:28:12gevaertstoffe82: the cable itself is 4-wire (ground, vbus, d+ and d-). The ID handling is purely in the connector, and isn't passed through to the other end
22:28:14amiconnbluebrother: AH, looks like I missed mentioning what to do at the special levels 'black' and 'white'. Those should be adjusted so that the central rectangle still looks black/white, even when looking from weird angles
22:28:24shotofaddswoohoo, I finally got the D2 to boot rockbox from flash rather than usb-boot :-)
22:28:43dionoeahi. Are any of the GSoC applicants for RockBox motivated to code MTP support? (just curious)
22:28:46Bagdermy D2 arrives any day now...
22:28:55linuxstbshotofadds: Nice ;)
22:28:57gevaertsdionoea: not yet
22:28:58shotofaddsBagder: you're getting quite a collection!
22:29:06linuxstbBagder: How many new targets are you buying?
22:29:13Bagderjust two
22:29:21amiconnOn most LCDs, that means just 2..5 steps away from the extreme. Gives better response for animated content when changing from white or black to some intermediate level
22:29:24linuxstbIs your D2 a DAB model?
22:29:39BagderI believe shotofadds has a DAB one
22:29:40shotofaddsunfortunately at this stage my hack involves putting a stragetically placed jump to the bootloader during the OF init code :/
22:29:50amiconnbluebrother: Your 'black' value looks pretty extreme, while the 'white' value is at full 255
22:29:53shotofaddsi need to work out what crucial piece of initialisation we are missing
22:30:14shotofaddslinuxstb: you were suitably impressed by my pub DAB demo, then ;)
22:30:49markungevaerts, toffe82: in the Meizu M6SL firmware I see the following strings: "Usb type is USB_HOST." and "Usb type is USB_FUNC."
22:31:24*gevaerts proposes to wait until we can run code on this thing, and then just try...
22:31:42*shotofadds goes back and buries his head in the disassembler...
22:31:57markungevaerts: I'm working on it :)
22:32:01*gevaerts spots another application
22:32:03*amiconn wonders what other choice gevaerts was thinking of
22:32:10toffe82gevaerts: in the cable pin 4 is connected to 5 so ID is grounded
22:32:50*gevaerts was choosing between 'wait' and 'go for a beer'
22:32:57Bagderapp #22 arrived
22:33:19markuntoffe82: but this is the older M6 model, the newer one (which gevaerts and I have) has slightly different hardware
22:33:35ceclinhi everybody
22:33:36markunBagder: how much did we have last year?
22:33:40markunhi ceclin
22:33:49*ceclin submitted an application for GSoC for rockbox
22:33:52ceclinfingers crossed!
22:33:54toffe82markun: ok
22:34:04Bagderceclin: which one?
22:34:14ceclinRealaudio support
22:34:22markunceclin: nice!
22:34:26 Join DerPapst [0] (
22:35:11Bagderceclin: here's a good guideline for info we'd like you to include:
22:35:21BagderI can see some of that is missing atm
22:35:38roolkuamiconn: my extremes are 10 and 252
22:35:53ceclinmissing? uh oh. let me give that a look
22:35:57domonokyceclin: also interesting for codecs, do you have any expirience with fixed-point code ?
22:36:14gevaertsBagder: what are you missing ?
22:36:44ceclindomonoky, no i'm afraid i don't.
22:36:48Bagderlet me retract that comment, everything that seems to be in there
22:37:02*Bagder types weirdly though...
22:37:33*gevaerts hands Bagder a new keyboard with Adaptive DWIM Technology
22:37:52 Quit dabujo (Connection timed out)
22:38:04 Quit Mathiasdm ("Yuuw!")
22:38:18domonokyceclin: could you add something about your coding expirience, so we know you are up to the task ? :-)
22:39:32*bluebrother hopes someone will reveal details about the applications to non-mentors
22:39:52ceclindomonoky: sure thing. Would a link to some code be useful as well?
22:39:59Bagderbluebrother: it'll happen, at least some of it
22:40:00ceclini mean, code that i've worked on
22:40:10domonokyceclin: that surely helps..
22:40:12gevaertsceclin: if you can legally link to it, sure
22:41:25*gevaerts decides to leave technical comments to people who have actually looked at codecs before
22:41:46 Quit MethoS-- (Success)
22:42:53toffe82low_light gevaerts : we should contact this guy :)
22:43:40 Join krz [0] (
22:43:58 Quit sarixe ("(EE) Failed to load "quit" module")
22:44:30*amiconn is calculating the mr100 lcd curve
22:45:23 Quit mf0102 ("Verlassend")
22:48:13gevaertstoffe82: Did you also spot sqlite in there ? :)
22:49:17 Quit desowin ()
22:49:42toffe82gevaerts: check this
22:49:54*gevaerts checks again ;)
22:50:07*shotofadds goes on a forehead-slapping rampage :(
22:50:22*Bagder ducks
22:50:26shotofaddsseems Rockbox works fine from flash. it's the dual-boot mechanism that's causing trouble.
22:50:43shotofaddslooks like the D2 won't boot anything (OF either) if a key is pressed on power on
22:50:57shotofaddsdual-boot could be tricky on this machine :/
22:53:16bluebrothershotofadds: is there a hold button?
22:53:51shotofaddsbluebrother: yes, but it's on a rocker with the power button - so maybe not usable for dualboot. i'll have to do some experiments.
22:54:38roolkushotofadds: use the powerbutton itself (like on m:robe?)
22:55:08 Join bertrik [0] (
22:55:27*shotofadds goes off to check the m:robe manual... is it some kind of long-press mechanism?
22:56:18roolkushotofadds: or the inofficial M5 bootloader does it in a similar way (short power=rockbox, long power =OF)
22:56:51roolkushotofadds: with the m:robe you need to let go briefly and press again
22:57:00preglowanyone with a 5g ipod here that can verify that bass/treble adjustments work as they should?
22:57:38 Nick HellDragon is now known as hd (i=jd@unaffiliated/helldragon)
22:57:44*gevaerts has a 5g, but never really understood what bass/treble should do, so he's probably not the right person
22:57:46 Quit Llorean ("Leaving.")
22:57:53preglowgevaerts: well, they adjust bass/treble...
22:58:04shotofaddsi'm doing some testing with the hold switch - there's a short delay after poweron while it loads the OF from flash, which might be long enough to flick the switch.
22:58:49gevaertspreglow: I probably could have guessed that. But my ears are no good at spotting if it's correct
22:59:04preglowoh well
22:59:16preglowi'm pretty sure it works as it should
22:59:20preglowso i'll just close this bugger
22:59:21 Quit low_light (" ajax IRC Client")
23:00:23shotofaddsyay, that works! hold power to boot the OF, or hold power then flick hold to boot Rockbox.
23:00:28shotofaddsnot ideal, but it works.
23:01:43Nico_Ptoffe82, gevaerts: this golb thing could be a nice test use case for our sqlite3 port
23:01:57*gevaerts spots yet another application
23:02:23bertriksqlite3 port to rockbox? wow
23:02:38DerPapstheh.. now they're all comming in ffff^^
23:03:00Nico_PI knew there would be more today... students always wait until deadlines :)
23:03:04gevaertsbertrik: sure. The only problem is that we don't really have a use for it yet :)
23:03:09amiconnroolku, bluebrother: Want a patch, or shall I commit right away?
23:03:27roolkuamiconn: just commit it I think
23:03:39Nico_Pbertrik: check out FS #8852 :)
23:03:54amiconnAlso, is there a button on m:robe that can be used for combos, ie. used as a kind of 'shift' key?
23:04:01*bertrik checks
23:04:23amiconnroolku: The greyscale plugins were already enabled for mr100?
23:04:44roolkuamiconn: yes (if you mean the games and demos)
23:05:08amiconngreyscale.rock isn't, and that one doesn't compile because of lacking button definitions
23:05:27amiconnIt's not enabled by default, by I have it enabled for my builds
23:05:46roolkuamiconn: I have the defines here, but its said it. :)
23:06:20amiconnOkay, I'll commit what my patch contained, plus the linearisation table
23:07:23 Quit krz (Read error: 110 (Connection timed out))
23:07:29roolkuamiconn: sounds good
23:09:31*domonoky doesnt understand the postion extracing out of the absolute packet in m:robes buttond drivers work.. too much shifting for me :-)
23:09:57roolkuamiconn: I believe the actual buttons (i.e. not the 3 fake ones on the strip) should be combinable
23:10:20toffe82Nico_P: waiting for an address book that I can sync with thunderbird
23:10:35roolkuamiconn: display is used in a few places as shift
23:11:04amiconnroolku: test_scanrate is also missing button definitions, so if you have that, you could commit it too
23:11:25roolkuamiconn: I will
23:12:42roolkudoom is crashing after presing start at the startup image :(
23:13:02amiconnDo you have voice enabled?
23:13:30*domonoky understands this positon extraction now, i looked at the wrong packet description :-)
23:14:47***Saving seen data "./dancer.seen"
23:16:38 Quit domonoky (Read error: 104 (Connection reset by peer))
23:16:59roolkuamiconn: I did, but even after disabling it it still crashes
23:17:01 Part Buschel
23:17:09amiconnWhat kind of crash?
23:17:37roolkuamiconn: frozen with random red pixels on black display
23:18:10*amiconn needs to check what the m:robe backlight handler does
23:18:52 Quit Jon-Kha (Remote closed the connection)
23:19:24amiconnHmm, nothing special, just GPO32
23:19:35*amiconn wonders what the problem could be here
23:20:58 Quit mcuelenaere ()
23:22:33 Quit Lear ("ChatZilla 0.9.81 [Firefox 3.0b5/2008032620]")
23:23:00 Join krabador [0] (n=DarkHamm@
23:24:56amiconnroolku: Weird. Do the other greyscale plugins work without such effects? (cube (shaded mode), fire, greyscale, mandelbrot, plasma, jpeg, mpegplayer, zxbox)
23:25:44 Quit bertrik (Read error: 104 (Connection reset by peer))
23:26:47 Join Jon-Kha [0] (
23:26:52roolkuI haven't tried zxbox, but the others work fine
23:28:33 Quit rob-e ("Ex-Chat")
23:28:58n1samiconn: does it just crash with voice enabled or with the file loaded too?
23:29:38amiconnIt doesn't crash if voice is disabled. I have voice files on all my targets, so they're loaded afaik
23:32:21 Quit dabujo_ ("( :: NoNameScript 4.2 :: )")
23:39:22 Join Jon-Kha_ [0] (
23:39:30 Quit Jon-Kha (Read error: 131 (Connection reset by peer))
23:39:30 Nick Jon-Kha_ is now known as Jon-Kha (
23:40:46 Quit lee-qid (Read error: 110 (Connection timed out))
23:41:43 Quit n1s ()
23:42:05 Quit tvelocity_ (Remote closed the connection)
23:43:16 Join Zarggg [0] (
23:44:14 Quit Zarggg (Read error: 104 (Connection reset by peer))
23:44:31 Join Zarggg [0] (
23:47:41 Nick fxb__ is now known as fxb (
23:48:17 Quit davina ("GNU/Linux the free alternative to Windows")
23:52:03 Nick DerPapst is now known as n (
23:52:10 Nick n is now known as DerPapst (
23:53:30 Quit Jon-Kha (Remote closed the connection)
23:53:36 Part Piero_Scarufii

Previous day | Next day